Artificial Intelligence Nanodegree

Convolutional Neural Networks

Project: Write an Algorithm for a Dog Identification App


In this notebook, some template code has already been provided for you, and you will need to implement additional functionality to successfully complete this project. You will not need to modify the included code beyond what is requested. Sections that begin with '(IMPLEMENTATION)' in the header indicate that the following block of code will require additional functionality which you must provide. Instructions will be provided for each section, and the specifics of the implementation are marked in the code block with a 'TODO' statement. Please be sure to read the instructions carefully!

Note: Once you have completed all of the code implementations, you need to finalize your work by exporting the iPython Notebook as an HTML document. Before exporting the notebook to html, all of the code cells need to have been run so that reviewers can see the final implementation and output. You can then export the notebook by using the menu above and navigating to \n", "File -> Download as -> HTML (.html). Include the finished document along with this notebook as your submission.

In addition to implementing code, there will be questions that you must answer which relate to the project and your implementation. Each section where you will answer a question is preceded by a 'Question X' header. Carefully read each question and provide thorough answers in the following text boxes that begin with 'Answer:'. Your project submission will be evaluated based on your answers to each of the questions and the implementation you provide.

Note: Code and Markdown cells can be executed using the Shift + Enter keyboard shortcut. Markdown cells can be edited by double-clicking the cell to enter edit mode.

The rubric contains optional "Stand Out Suggestions" for enhancing the project beyond the minimum requirements. If you decide to pursue the "Stand Out Suggestions", you should include the code in this IPython notebook.


Why We're Here

In this notebook, you will make the first steps towards developing an algorithm that could be used as part of a mobile or web app. At the end of this project, your code will accept any user-supplied image as input. If a dog is detected in the image, it will provide an estimate of the dog's breed. If a human is detected, it will provide an estimate of the dog breed that is most resembling. The image below displays potential sample output of your finished project (... but we expect that each student's algorithm will behave differently!).

Sample Dog Output

In this real-world setting, you will need to piece together a series of models to perform different tasks; for instance, the algorithm that detects humans in an image will be different from the CNN that infers dog breed. There are many points of possible failure, and no perfect algorithm exists. Your imperfect solution will nonetheless create a fun user experience!

The Road Ahead

We break the notebook into separate steps. Feel free to use the links below to navigate the notebook.

  • Step 0: Import Datasets
  • Step 1: Detect Humans
  • Step 2: Detect Dogs
  • Step 3: Create a CNN to Classify Dog Breeds (from Scratch)
  • Step 4: Use a CNN to Classify Dog Breeds (using Transfer Learning)
  • Step 5: Create a CNN to Classify Dog Breeds (using Transfer Learning)
  • Step 6: Write your Algorithm
  • Step 7: Test Your Algorithm

Step 0: Import Datasets

Import Dog Dataset

In the code cell below, we import a dataset of dog images. We populate a few variables through the use of the load_files function from the scikit-learn library:

  • train_files, valid_files, test_files - numpy arrays containing file paths to images
  • train_targets, valid_targets, test_targets - numpy arrays containing onehot-encoded classification labels
  • dog_names - list of string-valued dog breed names for translating labels
In [1]:
from sklearn.datasets import load_files       
from keras.utils import np_utils
import numpy as np
from glob import glob

# define function to load train, test, and validation datasets
def load_dataset(path):
    data = load_files(path)
    dog_files = np.array(data['filenames'])
    dog_targets = np_utils.to_categorical(np.array(data['target']), 133)
    return dog_files, dog_targets
dogImages ='C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/dogImages'
# load train, test, and validation datasets
#train_files, train_targets = load_dataset('dogImages/train')
#valid_files, valid_targets = load_dataset('dogImages/valid')
#test_files, test_targets = load_dataset('dogImages/test')
train_files, train_targets = load_dataset(dogImages+'/train')
valid_files, valid_targets = load_dataset(dogImages+'/valid')
test_files, test_targets = load_dataset(dogImages+'/test')

# load list of dog names
#dog_names = [item[20:-1] for item in sorted(glob("dogImages/train/*/"))]
dog_names = [item[108:-1] for item in glob(dogImages+'/train/*/')]
#print(dog_names,'\n')

# print statistics about the dataset
print('There are %d total dog categories.' % len(dog_names))
print('There are %s total dog images.\n' % len(np.hstack([train_files, valid_files, test_files])))
print('There are %d training dog images.' % len(train_files))
print('There are %d validation dog images.' % len(valid_files))
print('There are %d test dog images.'% len(test_files))
Using TensorFlow backend.
There are 133 total dog categories.
There are 8351 total dog images.

There are 6680 training dog images.
There are 835 validation dog images.
There are 836 test dog images.

Import Human Dataset

In the code cell below, we import a dataset of human images, where the file paths are stored in the numpy array human_files.

In [2]:
import random
random.seed(8675309)
datadir ='C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/'
# load filenames in shuffled human dataset
#human_files = np.array(glob(datadir+"lfw/*/*"))
human_files = np.array(glob(datadir+'lfw/lfw/*/*'))
random.shuffle(human_files)
print(human_files[3])

# print statistics about the dataset
print('There are %d total human images.' % len(human_files))
C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/lfw/lfw\Laurence_Fishburne\Laurence_Fishburne_0001.jpg
There are 13233 total human images.
In [3]:
def load_human_dataset(path):    
    data = load_files(path)
    human_files1 = np.array(data['filenames'])
    human_targets1 = np_utils.to_categorical(np.array(data['target']), 13233)
    return human_files1, human_targets1

human_files1, human_targets1 = load_human_dataset(datadir+'lfw/lfw/')

print(human_files1[3])
print(human_targets1[3])
C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/lfw/lfw/Tang_Jiaxuan\Tang_Jiaxuan_0009.jpg
[ 0.  0.  0. ...,  0.  0.  0.]

Step 1: Detect Humans

We use OpenCV's implementation of Haar feature-based cascade classifiers to detect human faces in images. OpenCV provides many pre-trained face detectors, stored as XML files on github. We have downloaded one of these detectors and stored it in the haarcascades directory.

In the next code cell, we demonstrate how to use this detector to find human faces in a sample image.

In [4]:
import cv2                
import matplotlib.pyplot as plt                        
%matplotlib inline                               

# extract pre-trained face detector

#face_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_frontalface_alt.xml')
face_cascade = cv2.CascadeClassifier(datadir+'opencv-master/data/'+'haarcascades/haarcascade_frontalface_alt.xml')
#print(face_cascade)
# load color (BGR) image
random.shuffle(human_files)
img = cv2.imread(human_files[3])
print("\n",human_files[3])
# convert BGR image to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

# find faces in image
faces = face_cascade.detectMultiScale(gray)

# print number of faces detected in the image
print('Number of faces detected:', len(faces))

# get bounding box for each detected face
for (x,y,w,h) in faces:
    # add bounding box to color image#
    cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
    
# convert BGR image to RGB for plotting
cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)

# display the image, along with bounding box
plt.imshow(cv_rgb)
plt.show()
 C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/lfw/lfw\Casey_Mears\Casey_Mears_0001.jpg
Number of faces detected: 1

Before using any of the face detectors, it is standard procedure to convert the images to grayscale. The detectMultiScale function executes the classifier stored in face_cascade and takes the grayscale image as a parameter.

In the above code, faces is a numpy array of detected faces, where each row corresponds to a detected face. Each detected face is a 1D array with four entries that specifies the bounding box of the detected face. The first two entries in the array (extracted in the above code as x and y) specify the horizontal and vertical positions of the top left corner of the bounding box. The last two entries in the array (extracted here as w and h) specify the width and height of the box.

Write a Human Face Detector

We can use this procedure to write a function that returns True if a human face is detected in an image and False otherwise. This function, aptly named face_detector, takes a string-valued file path to an image as input and appears in the code block below.

In [5]:
# returns "True" if face is detected in image stored at img_path
def face_detector(img_path):
    img = cv2.imread(img_path)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
    faces = face_cascade.detectMultiScale(gray)
    return len(faces) > 0

(IMPLEMENTATION) Assess the Human Face Detector

Question 1: Use the code cell below to test the performance of the face_detector function.

  • What percentage of the first 100 images in human_files have a detected human face?
  • What percentage of the first 100 images in dog_files have a detected human face?

Ideally, we would like 100% of human images with a detected face and 0% of dog images with a detected face. You will see that our algorithm falls short of this goal, but still gives acceptable performance. We extract the file paths for the first 100 images from each of the datasets and store them in the numpy arrays human_files_short and dog_files_short.

Answer:
What percentage of the first 100 images in human_files have a detected human face? = 99%
What percentage of the first 100 images in dog_files have a detected human face? = 11%

In [6]:
human_files_short = human_files[:100]
dog_files_short = train_files[:100]
# Do NOT modify the code above this line.

## TODO: Test the performance of the face_detector algorithm 
## on the images in human_files_short and dog_files_short.
human_file_count =0
for human_file in human_files_short:
    if face_detector(human_file):
        human_file_count +=1
print('Human Face Detected out of %d  files is %d  ' % (len(human_files_short),  human_file_count))        

dog_file_count =0
for dog_file in dog_files_short:
    if face_detector(dog_file):
        dog_file_count +=1
print('Human Face Detected in %d Dog files is %d  ' % (len(dog_files_short),  dog_file_count))
Human Face Detected out of 100  files is 99  
Human Face Detected in 100 Dog files is 11  

Question 2: This algorithmic choice necessitates that we communicate to the user that we accept human images only when they provide a clear view of a face (otherwise, we risk having unneccessarily frustrated users!). In your opinion, is this a reasonable expectation to pose on the user? If not, can you think of a way to detect humans in images that does not necessitate an image with a clearly presented face?

Answer:
This is not a reasonable exceptation. Use data agumentation logic to detect humans in images that does not necessitate an image with a clearly presented face.


Example: Simple use of ImageDataGenerator for data agumentation. This sample shows rotation of image by 30 degree, width and height sift by 20% Shear and Zoom by 20% . Flipping the Image horizontally.


ImageDataGenerator(preprocessing_function=preprocess_input,
rotation_range=30,
width_shift_range=0.2,
height_shift_range=0.2,
shear_range=0.2,
zoom_range=0.2,
horizontal_flip=True )


Here are the other list of parameters that you can apply to agument images:

  • featurewise_center: Boolean. Set input mean to 0 over the dataset, feature-wise.

  • samplewise_center: Boolean. Set each sample mean to 0.

  • featurewise_std_normalization: Boolean. Divide inputs by std of the dataset, feature-wise.

  • samplewise_std_normalization: Boolean. Divide each input by its std.

  • zca_epsilon: epsilon for ZCA whitening. Default is 1e-6.

  • zca_whitening: Boolean. Apply ZCA whitening.

  • rotation_range: Int. Degree range for random rotations.

  • width_shift_range: Float (fraction of total width). Range for random horizontal shifts.

  • height_shift_range: Float (fraction of total height). Range for random vertical shifts.

  • shear_range: Float. Shear Intensity (Shear angle in counter-clockwise direction as radians)

  • zoom_range: Float or [lower, upper]. Range for random zoom. If a float, [lower, upper] = [1-zoom_range, 1+zoom_range].

  • channel_shift_range: Float. Range for random channel shifts.

  • fill_mode: One of {"constant", "nearest", "reflect" or "wrap"}. Points outside the boundaries of the input are filled according to the given mode.

  • cval: Float or Int. Value used for points outside the boundaries when fill_mode = "constant".

  • horizontal_flip: Boolean. Randomly flip inputs horizontally.

  • vertical_flip: Boolean. Randomly flip inputs vertically.

  • rescale: rescaling factor. Defaults to None. If None or 0, no rescaling is applied, otherwise we multiply the data by the value provided (before applying any other transformation).

  • preprocessing_function: function that will be implied on each input. The function will run before any other modification on it. The function should take one argument: one image (Numpy tensor with rank 3), and should output a Numpy tensor with the same shape.

  • data_format: One of {"channels_first", "channels_last"}. "channels_last" mode means that the images should have shape (samples, height, width, channels), "channels_first" mode means that the images should have shape (samples, channels, height, width). It defaults to the image_data_format value found in your Keras config file at ~/.keras/keras.json. If you never set it, then it will be "channels_last"


    We suggest the face detector from OpenCV as a potential way to detect human images in your algorithm, but you are free to explore other approaches, especially approaches that make use of deep learning :). Please use the code cell below to design and test your own face detection algorithm. If you decide to pursue this optional task, report performance on each of the datasets.

  • In [124]:
    ## (Optional) TODO: Report the performance of another  
    ## face detection algorithm on the LFW dataset
    ### Feel free to use as many code cells as needed.
    
    #print(" Before Removing Final Layer \n",model.predict(human_files).shape)
    #from keras.applications.vgg16 import VGG16
    #model = VGG16(include_top=False)
    #model.summary()
    
    #print(" After Removing Final Layer \n",model.predict(human_files).shape)
    
    face_cascade1 = cv2.CascadeClassifier(datadir+'opencv-master/data/'+'haarcascades/haarcascade_frontalface_default.xml')
    random.shuffle(human_files)
    img1 = cv2.imread(human_files[3])
    print("\n",human_files[3][109:-4])
    # convert BGR image to grayscale
    gray1 = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
    
    
    # find faces in image
    faces1 = face_cascade1.detectMultiScale(gray1,scaleFactor=1.1,
        minNeighbors=5,
        minSize=(20, 20)
        )
    
    # print number of faces detected in the image
    print('Number of faces detected:', len(faces1))
    
    # get bounding box for each detected face
    for (x,y,w,h) in faces1:
        # add bounding box to color image#
        cv2.rectangle(img1,(x,y),(x+w,y+h),(255,0,0),2)
        
    # convert BGR image to RGB for plotting
    cv_rgb1 = cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)
    
    # display the image, along with bounding box
    plt.imshow(cv_rgb1)
    plt.show()
    
     d\Ernie_Grunfeld_0001
    Number of faces detected: 1
    
    In [125]:
    def face_detector1(img_path):
        img1 = cv2.imread(img_path)
        gray1 = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
        faces1 = face_cascade1.detectMultiScale(gray1,scaleFactor=1.1,
        minNeighbors=5,
        minSize=(20, 20)
        )
        return len(faces1) > 0
    
    In [126]:
    human_files_short1 = human_files[:100]
    dog_files_short1 = train_files[:100]
    # Do NOT modify the code above this line.
    
    ## TODO: Test the performance of the face_detector algorithm 
    ## on the images in human_files_short and dog_files_short.
    human_file_count1 =0
    for human_file1 in human_files_short1:
        if face_detector1(human_file1):
            human_file_count1 +=1
    print('Human Face Detected out of %d  files is %d  ' % (len(human_files_short1),  human_file_count1))        
    
    dog_file_count1 =0
    for dog_file1 in dog_files_short1:
        if face_detector1(dog_file1):
            dog_file_count1 +=1
    print('Human Face Detected in %d Dog files is %d  ' % (len(dog_files_short1),  dog_file_count1))
    
    Human Face Detected out of 100  files is 99  
    Human Face Detected in 100 Dog files is 33  
    

    Step 2: Detect Dogs

    In this section, we use a pre-trained ResNet-50 model to detect dogs in images. Our first line of code downloads the ResNet-50 model, along with weights that have been trained on ImageNet, a very large, very popular dataset used for image classification and other vision tasks. ImageNet contains over 10 million URLs, each linking to an image containing an object from one of 1000 categories. Given an image, this pre-trained ResNet-50 model returns a prediction (derived from the available categories in ImageNet) for the object that is contained in the image.

    In [7]:
    from keras.applications.resnet50 import ResNet50
    
    # define ResNet50 model
    ResNet50_model = ResNet50(weights='imagenet')
    
    ResNet50_model.summary()
    
    ____________________________________________________________________________________________________
    Layer (type)                     Output Shape          Param #     Connected to                     
    ====================================================================================================
    input_1 (InputLayer)             (None, 224, 224, 3)   0                                            
    ____________________________________________________________________________________________________
    zero_padding2d_1 (ZeroPadding2D) (None, 230, 230, 3)   0                                            
    ____________________________________________________________________________________________________
    conv1 (Conv2D)                   (None, 112, 112, 64)  9472                                         
    ____________________________________________________________________________________________________
    bn_conv1 (BatchNormalization)    (None, 112, 112, 64)  256                                          
    ____________________________________________________________________________________________________
    activation_1 (Activation)        (None, 112, 112, 64)  0                                            
    ____________________________________________________________________________________________________
    max_pooling2d_1 (MaxPooling2D)   (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2a_branch2a (Conv2D)          (None, 55, 55, 64)    4160                                         
    ____________________________________________________________________________________________________
    bn2a_branch2a (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_2 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2a_branch2b (Conv2D)          (None, 55, 55, 64)    36928                                        
    ____________________________________________________________________________________________________
    bn2a_branch2b (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_3 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2a_branch2c (Conv2D)          (None, 55, 55, 256)   16640                                        
    ____________________________________________________________________________________________________
    res2a_branch1 (Conv2D)           (None, 55, 55, 256)   16640                                        
    ____________________________________________________________________________________________________
    bn2a_branch2c (BatchNormalizatio (None, 55, 55, 256)   1024                                         
    ____________________________________________________________________________________________________
    bn2a_branch1 (BatchNormalization (None, 55, 55, 256)   1024                                         
    ____________________________________________________________________________________________________
    add_1 (Add)                      (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    activation_4 (Activation)        (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    res2b_branch2a (Conv2D)          (None, 55, 55, 64)    16448                                        
    ____________________________________________________________________________________________________
    bn2b_branch2a (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_5 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2b_branch2b (Conv2D)          (None, 55, 55, 64)    36928                                        
    ____________________________________________________________________________________________________
    bn2b_branch2b (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_6 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2b_branch2c (Conv2D)          (None, 55, 55, 256)   16640                                        
    ____________________________________________________________________________________________________
    bn2b_branch2c (BatchNormalizatio (None, 55, 55, 256)   1024                                         
    ____________________________________________________________________________________________________
    add_2 (Add)                      (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    activation_7 (Activation)        (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    res2c_branch2a (Conv2D)          (None, 55, 55, 64)    16448                                        
    ____________________________________________________________________________________________________
    bn2c_branch2a (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_8 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2c_branch2b (Conv2D)          (None, 55, 55, 64)    36928                                        
    ____________________________________________________________________________________________________
    bn2c_branch2b (BatchNormalizatio (None, 55, 55, 64)    256                                          
    ____________________________________________________________________________________________________
    activation_9 (Activation)        (None, 55, 55, 64)    0                                            
    ____________________________________________________________________________________________________
    res2c_branch2c (Conv2D)          (None, 55, 55, 256)   16640                                        
    ____________________________________________________________________________________________________
    bn2c_branch2c (BatchNormalizatio (None, 55, 55, 256)   1024                                         
    ____________________________________________________________________________________________________
    add_3 (Add)                      (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    activation_10 (Activation)       (None, 55, 55, 256)   0                                            
    ____________________________________________________________________________________________________
    res3a_branch2a (Conv2D)          (None, 28, 28, 128)   32896                                        
    ____________________________________________________________________________________________________
    bn3a_branch2a (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_11 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3a_branch2b (Conv2D)          (None, 28, 28, 128)   147584                                       
    ____________________________________________________________________________________________________
    bn3a_branch2b (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_12 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3a_branch2c (Conv2D)          (None, 28, 28, 512)   66048                                        
    ____________________________________________________________________________________________________
    res3a_branch1 (Conv2D)           (None, 28, 28, 512)   131584                                       
    ____________________________________________________________________________________________________
    bn3a_branch2c (BatchNormalizatio (None, 28, 28, 512)   2048                                         
    ____________________________________________________________________________________________________
    bn3a_branch1 (BatchNormalization (None, 28, 28, 512)   2048                                         
    ____________________________________________________________________________________________________
    add_4 (Add)                      (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    activation_13 (Activation)       (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    res3b_branch2a (Conv2D)          (None, 28, 28, 128)   65664                                        
    ____________________________________________________________________________________________________
    bn3b_branch2a (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_14 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3b_branch2b (Conv2D)          (None, 28, 28, 128)   147584                                       
    ____________________________________________________________________________________________________
    bn3b_branch2b (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_15 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3b_branch2c (Conv2D)          (None, 28, 28, 512)   66048                                        
    ____________________________________________________________________________________________________
    bn3b_branch2c (BatchNormalizatio (None, 28, 28, 512)   2048                                         
    ____________________________________________________________________________________________________
    add_5 (Add)                      (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    activation_16 (Activation)       (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    res3c_branch2a (Conv2D)          (None, 28, 28, 128)   65664                                        
    ____________________________________________________________________________________________________
    bn3c_branch2a (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_17 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3c_branch2b (Conv2D)          (None, 28, 28, 128)   147584                                       
    ____________________________________________________________________________________________________
    bn3c_branch2b (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_18 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3c_branch2c (Conv2D)          (None, 28, 28, 512)   66048                                        
    ____________________________________________________________________________________________________
    bn3c_branch2c (BatchNormalizatio (None, 28, 28, 512)   2048                                         
    ____________________________________________________________________________________________________
    add_6 (Add)                      (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    activation_19 (Activation)       (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    res3d_branch2a (Conv2D)          (None, 28, 28, 128)   65664                                        
    ____________________________________________________________________________________________________
    bn3d_branch2a (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_20 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3d_branch2b (Conv2D)          (None, 28, 28, 128)   147584                                       
    ____________________________________________________________________________________________________
    bn3d_branch2b (BatchNormalizatio (None, 28, 28, 128)   512                                          
    ____________________________________________________________________________________________________
    activation_21 (Activation)       (None, 28, 28, 128)   0                                            
    ____________________________________________________________________________________________________
    res3d_branch2c (Conv2D)          (None, 28, 28, 512)   66048                                        
    ____________________________________________________________________________________________________
    bn3d_branch2c (BatchNormalizatio (None, 28, 28, 512)   2048                                         
    ____________________________________________________________________________________________________
    add_7 (Add)                      (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    activation_22 (Activation)       (None, 28, 28, 512)   0                                            
    ____________________________________________________________________________________________________
    res4a_branch2a (Conv2D)          (None, 14, 14, 256)   131328                                       
    ____________________________________________________________________________________________________
    bn4a_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_23 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4a_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4a_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_24 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4a_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    res4a_branch1 (Conv2D)           (None, 14, 14, 1024)  525312                                       
    ____________________________________________________________________________________________________
    bn4a_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    bn4a_branch1 (BatchNormalization (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_8 (Add)                      (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_25 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res4b_branch2a (Conv2D)          (None, 14, 14, 256)   262400                                       
    ____________________________________________________________________________________________________
    bn4b_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_26 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4b_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4b_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_27 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4b_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    bn4b_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_9 (Add)                      (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_28 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res4c_branch2a (Conv2D)          (None, 14, 14, 256)   262400                                       
    ____________________________________________________________________________________________________
    bn4c_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_29 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4c_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4c_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_30 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4c_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    bn4c_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_10 (Add)                     (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_31 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res4d_branch2a (Conv2D)          (None, 14, 14, 256)   262400                                       
    ____________________________________________________________________________________________________
    bn4d_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_32 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4d_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4d_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_33 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4d_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    bn4d_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_11 (Add)                     (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_34 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res4e_branch2a (Conv2D)          (None, 14, 14, 256)   262400                                       
    ____________________________________________________________________________________________________
    bn4e_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_35 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4e_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4e_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_36 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4e_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    bn4e_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_12 (Add)                     (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_37 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res4f_branch2a (Conv2D)          (None, 14, 14, 256)   262400                                       
    ____________________________________________________________________________________________________
    bn4f_branch2a (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_38 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4f_branch2b (Conv2D)          (None, 14, 14, 256)   590080                                       
    ____________________________________________________________________________________________________
    bn4f_branch2b (BatchNormalizatio (None, 14, 14, 256)   1024                                         
    ____________________________________________________________________________________________________
    activation_39 (Activation)       (None, 14, 14, 256)   0                                            
    ____________________________________________________________________________________________________
    res4f_branch2c (Conv2D)          (None, 14, 14, 1024)  263168                                       
    ____________________________________________________________________________________________________
    bn4f_branch2c (BatchNormalizatio (None, 14, 14, 1024)  4096                                         
    ____________________________________________________________________________________________________
    add_13 (Add)                     (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    activation_40 (Activation)       (None, 14, 14, 1024)  0                                            
    ____________________________________________________________________________________________________
    res5a_branch2a (Conv2D)          (None, 7, 7, 512)     524800                                       
    ____________________________________________________________________________________________________
    bn5a_branch2a (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_41 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5a_branch2b (Conv2D)          (None, 7, 7, 512)     2359808                                      
    ____________________________________________________________________________________________________
    bn5a_branch2b (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_42 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5a_branch2c (Conv2D)          (None, 7, 7, 2048)    1050624                                      
    ____________________________________________________________________________________________________
    res5a_branch1 (Conv2D)           (None, 7, 7, 2048)    2099200                                      
    ____________________________________________________________________________________________________
    bn5a_branch2c (BatchNormalizatio (None, 7, 7, 2048)    8192                                         
    ____________________________________________________________________________________________________
    bn5a_branch1 (BatchNormalization (None, 7, 7, 2048)    8192                                         
    ____________________________________________________________________________________________________
    add_14 (Add)                     (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    activation_43 (Activation)       (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    res5b_branch2a (Conv2D)          (None, 7, 7, 512)     1049088                                      
    ____________________________________________________________________________________________________
    bn5b_branch2a (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_44 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5b_branch2b (Conv2D)          (None, 7, 7, 512)     2359808                                      
    ____________________________________________________________________________________________________
    bn5b_branch2b (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_45 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5b_branch2c (Conv2D)          (None, 7, 7, 2048)    1050624                                      
    ____________________________________________________________________________________________________
    bn5b_branch2c (BatchNormalizatio (None, 7, 7, 2048)    8192                                         
    ____________________________________________________________________________________________________
    add_15 (Add)                     (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    activation_46 (Activation)       (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    res5c_branch2a (Conv2D)          (None, 7, 7, 512)     1049088                                      
    ____________________________________________________________________________________________________
    bn5c_branch2a (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_47 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5c_branch2b (Conv2D)          (None, 7, 7, 512)     2359808                                      
    ____________________________________________________________________________________________________
    bn5c_branch2b (BatchNormalizatio (None, 7, 7, 512)     2048                                         
    ____________________________________________________________________________________________________
    activation_48 (Activation)       (None, 7, 7, 512)     0                                            
    ____________________________________________________________________________________________________
    res5c_branch2c (Conv2D)          (None, 7, 7, 2048)    1050624                                      
    ____________________________________________________________________________________________________
    bn5c_branch2c (BatchNormalizatio (None, 7, 7, 2048)    8192                                         
    ____________________________________________________________________________________________________
    add_16 (Add)                     (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    activation_49 (Activation)       (None, 7, 7, 2048)    0                                            
    ____________________________________________________________________________________________________
    avg_pool (AveragePooling2D)      (None, 1, 1, 2048)    0                                            
    ____________________________________________________________________________________________________
    flatten_1 (Flatten)              (None, 2048)          0                                            
    ____________________________________________________________________________________________________
    fc1000 (Dense)                   (None, 1000)          2049000                                      
    ====================================================================================================
    Total params: 25,636,712.0
    Trainable params: 25,583,592.0
    Non-trainable params: 53,120.0
    ____________________________________________________________________________________________________
    

    Pre-process the Data

    When using TensorFlow as backend, Keras CNNs require a 4D array (which we'll also refer to as a 4D tensor) as input, with shape

    $$ (\text{nb_samples}, \text{rows}, \text{columns}, \text{channels}), $$

    where nb_samples corresponds to the total number of images (or samples), and rows, columns, and channels correspond to the number of rows, columns, and channels for each image, respectively.

    The path_to_tensor function below takes a string-valued file path to a color image as input and returns a 4D tensor suitable for supplying to a Keras CNN. The function first loads the image and resizes it to a square image that is $224 \times 224$ pixels. Next, the image is converted to an array, which is then resized to a 4D tensor. In this case, since we are working with color images, each image has three channels. Likewise, since we are processing a single image (or sample), the returned tensor will always have shape

    $$ (1, 224, 224, 3). $$

    The paths_to_tensor function takes a numpy array of string-valued image paths as input and returns a 4D tensor with shape

    $$ (\text{nb_samples}, 224, 224, 3). $$

    Here, nb_samples is the number of samples, or number of images, in the supplied array of image paths. It is best to think of nb_samples as the number of 3D tensors (where each 3D tensor corresponds to a different image) in your dataset!

    In [8]:
    from keras.preprocessing import image                  
    from tqdm import tqdm
    
    def path_to_tensor(img_path):
        #print(img_path)
        # loads RGB image as PIL.Image.Image type
        img = image.load_img(img_path, target_size=(224, 224))
        # convert PIL.Image.Image type to 3D tensor with shape (224, 224, 3)
        x = image.img_to_array(img)
        # convert 3D tensor to 4D tensor with shape (1, 224, 224, 3) and return 4D tensor
        return np.expand_dims(x, axis=0)
    
    def paths_to_tensor(img_paths):
        list_of_tensors = [path_to_tensor(img_path) for img_path in tqdm(img_paths)]
        return np.vstack(list_of_tensors)
    

    Making Predictions with ResNet-50

    Getting the 4D tensor ready for ResNet-50, and for any other pre-trained model in Keras, requires some additional processing. First, the RGB image is converted to BGR by reordering the channels. All pre-trained models have the additional normalization step that the mean pixel (expressed in RGB as $[103.939, 116.779, 123.68]$ and calculated from all pixels in all images in ImageNet) must be subtracted from every pixel in each image. This is implemented in the imported function preprocess_input. If you're curious, you can check the code for preprocess_input here.

    Now that we have a way to format our image for supplying to ResNet-50, we are now ready to use the model to extract the predictions. This is accomplished with the predict method, which returns an array whose $i$-th entry is the model's predicted probability that the image belongs to the $i$-th ImageNet category. This is implemented in the ResNet50_predict_labels function below.

    By taking the argmax of the predicted probability vector, we obtain an integer corresponding to the model's predicted object class, which we can identify with an object category through the use of this dictionary.

    In [9]:
    from keras.applications.resnet50 import preprocess_input, decode_predictions
    
    def ResNet50_predict_labels(img_path):
        # returns prediction vector for image located at img_path
        img = preprocess_input(path_to_tensor(img_path))
        return np.argmax(ResNet50_model.predict(img))
    

    Write a Dog Detector

    While looking at the dictionary, you will notice that the categories corresponding to dogs appear in an uninterrupted sequence and correspond to dictionary keys 151-268, inclusive, to include all categories from 'Chihuahua' to 'Mexican hairless'. Thus, in order to check to see if an image is predicted to contain a dog by the pre-trained ResNet-50 model, we need only check if the ResNet50_predict_labels function above returns a value between 151 and 268 (inclusive).

    We use these ideas to complete the dog_detector function below, which returns True if a dog is detected in an image (and False if not).

    In [10]:
    ### returns "True" if a dog is detected in the image stored at img_path
    def dog_detector(img_path):
        prediction = ResNet50_predict_labels(img_path)
        return ((prediction <= 268) & (prediction >= 151)) 
    

    (IMPLEMENTATION) Assess the Dog Detector

    Question 3: Use the code cell below to test the performance of your dog_detector function.

    • What percentage of the images in human_files_short have a detected dog?
    • What percentage of the images in dog_files_short have a detected dog?

    Answer:
    Human file detected as dog is 1%
    Dog file detected as dog is 100%

    In [11]:
    ### TODO: Test the performance of the dog_detector function
    ### on the images in human_files_short and dog_files_short.
    
    human_file_dog_count = 0
    for human_file in human_files_short:
        if dog_detector(human_file):
            human_file_dog_count +=  1
            
    dog_file_dog_count = 0
    for dog_file in dog_files_short:
        if dog_detector(dog_file):
            dog_file_dog_count +=  1
    #print('Human Face Detected out of %d  files is %d  ' % (len(human_files_short1),  human_file_count1))        
            
    print('Human file detected as dog is %s ' % str(human_file_dog_count) )
    print('Dog file detected as dog is %s ' % str(dog_file_dog_count) )
    
    Human file detected as dog is 1 
    Dog file detected as dog is 100 
    

    Step 3: Create a CNN to Classify Dog Breeds (from Scratch)

    Now that we have functions for detecting humans and dogs in images, we need a way to predict breed from images. In this step, you will create a CNN that classifies dog breeds. You must create your CNN from scratch (so, you can't use transfer learning yet!), and you must attain a test accuracy of at least 1%. In Step 5 of this notebook, you will have the opportunity to use transfer learning to create a CNN that attains greatly improved accuracy.

    Be careful with adding too many trainable layers! More parameters means longer training, which means you are more likely to need a GPU to accelerate the training process. Thankfully, Keras provides a handy estimate of the time that each epoch is likely to take; you can extrapolate this estimate to figure out how long it will take for your algorithm to train.

    We mention that the task of assigning breed to dogs from images is considered exceptionally challenging. To see why, consider that even a human would have great difficulty in distinguishing between a Brittany and a Welsh Springer Spaniel.

    Brittany Welsh Springer Spaniel

    It is not difficult to find other dog breed pairs with minimal inter-class variation (for instance, Curly-Coated Retrievers and American Water Spaniels).

    Curly-Coated Retriever American Water Spaniel

    Likewise, recall that labradors come in yellow, chocolate, and black. Your vision-based algorithm will have to conquer this high intra-class variation to determine how to classify all of these different shades as the same breed.

    Yellow Labrador Chocolate Labrador Black Labrador

    We also mention that random chance presents an exceptionally low bar: setting aside the fact that the classes are slightly imabalanced, a random guess will provide a correct answer roughly 1 in 133 times, which corresponds to an accuracy of less than 1%.

    Remember that the practice is far ahead of the theory in deep learning. Experiment with many different architectures, and trust your intuition. And, of course, have fun!

    Pre-process the Data

    We rescale the images by dividing every pixel in every image by 255.

    In [12]:
    from PIL import ImageFile                            
    ImageFile.LOAD_TRUNCATED_IMAGES = True                 
    
    # pre-process the data for Keras
    train_tensors = paths_to_tensor(train_files).astype('float32')/255
    valid_tensors = paths_to_tensor(valid_files).astype('float32')/255
    test_tensors = paths_to_tensor(test_files).astype('float32')/255
    
    100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 6680/6680 [01:47<00:00, 61.92it/s]
    100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 835/835 [01:19<00:00, 10.56it/s]
    100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 836/836 [00:43<00:00, 19.02it/s]
    

    (IMPLEMENTATION) Model Architecture

    Create a CNN to classify dog breed. At the end of your code cell block, summarize the layers of your model by executing the line:

        model.summary()
    
    

    We have imported some Python modules to get you started, but feel free to import as many modules as you need. If you end up getting stuck, here's a hint that specifies a model that trains relatively fast on CPU and attains >1% test accuracy in 5 epochs:

    Sample CNN

    Question 4: Outline the steps you took to get to your final CNN architecture and your reasoning at each step. If you chose to use the hinted architecture above, describe why you think that CNN architecture should work well for the image classification task.

    Answer:
    CNN is a good fit for image classification, it uses deeplearning to process each lay and use it in the next layer to further enhance it. In case of CNN in the First layer edges and lines are identified in the next layer it uses these lines and edges to identify the next step that is the shapes and so on and finally the last layer identifies the details of object which in this case the eyes , teeth , face of the dog or human.

  • The final CNN architecture consists of 3 Convolution followed by 3 Max pool layers

  • The first 6 layers are designed to take input array of image pixels and convert to an array where all spatial info is squeezed out and only info encoding the content of the image remains.

  • Conv2D starts with a deapth of 16 and gradually the deapth is increased to 64 and the MaxPool with pool size of 2 is cutting down the shape by half each time (224 to 112 to 56 to 28 and finally to 7) where I introduce the Global Average pooling to further elucidate the content of image .

  • The final layer is a dense layer and has one entry for each object class in the dataset (in this case we have 133 object class).

  • The Softmax activation fuction in the final layer returns the probablity.

  • In [13]:
    from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D
    from keras.layers import Dropout, Flatten, Dense
    from keras.models import Sequential
    
    model = Sequential()
    
    
    ### TODO: Define your architecture.
    model.add(Conv2D(filters=16, kernel_size=2, padding='same', activation='relu', input_shape=(224, 224, 3)))
    model.add(MaxPooling2D(pool_size=2))
    model.add(Conv2D(filters=32, kernel_size=2, padding='same', activation='relu'))
    model.add(MaxPooling2D(pool_size=2))
    model.add(Conv2D(filters=64, kernel_size=2, padding='same', activation='relu'))
    model.add(MaxPooling2D(pool_size=2))
    model.add(GlobalAveragePooling2D(input_shape=(7, 7, 512)))
    model.add(Dense(133, activation='softmax'))
    model.summary()
    print('Hello')
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    conv2d_1 (Conv2D)            (None, 224, 224, 16)      208       
    _________________________________________________________________
    max_pooling2d_2 (MaxPooling2 (None, 112, 112, 16)      0         
    _________________________________________________________________
    conv2d_2 (Conv2D)            (None, 112, 112, 32)      2080      
    _________________________________________________________________
    max_pooling2d_3 (MaxPooling2 (None, 56, 56, 32)        0         
    _________________________________________________________________
    conv2d_3 (Conv2D)            (None, 56, 56, 64)        8256      
    _________________________________________________________________
    max_pooling2d_4 (MaxPooling2 (None, 28, 28, 64)        0         
    _________________________________________________________________
    global_average_pooling2d_1 ( (None, 64)                0         
    _________________________________________________________________
    dense_1 (Dense)              (None, 133)               8645      
    =================================================================
    Total params: 19,189.0
    Trainable params: 19,189.0
    Non-trainable params: 0.0
    _________________________________________________________________
    Hello
    

    Compile the Model

    In [14]:
    model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])
    

    (IMPLEMENTATION) Train the Model

    Train your model in the code cell below. Use model checkpointing to save the model that attains the best validation loss.

    You are welcome to augment the training data, but this is not a requirement.

    In [15]:
    from keras.callbacks import ModelCheckpoint  
    
    ### TODO: specify the number of epochs that you would like to use to train the model.
    
    epochs = 10
    
    ### Do NOT modify the code below this line.
    
    checkpointer = ModelCheckpoint(filepath='weights.best.from_scratch.hdf5', 
                                   verbose=1, save_best_only=True)
    
    model.fit(train_tensors, train_targets, 
              validation_data=(valid_tensors, valid_targets),
              epochs=epochs, batch_size=20, callbacks=[checkpointer], verbose=1)
    
    print('--- Sction Complete ----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/10
    6660/6680 [============================>.] - ETA: 602s - loss: 4.8835 - acc: 0.0000e+00 - ETA: 438s - loss: 4.8868 - acc: 0.0000e+00 - ETA: 378s - loss: 4.8901 - acc: 0.0000e+00 - ETA: 349s - loss: 4.8926 - acc: 0.0000e+00 - ETA: 334s - loss: 4.8928 - acc: 0.0000e+00 - ETA: 321s - loss: 4.8917 - acc: 0.0000e+00 - ETA: 314s - loss: 4.8916 - acc: 0.0000e+00 - ETA: 309s - loss: 4.8918 - acc: 0.0000e+00 - ETA: 303s - loss: 4.8912 - acc: 0.0000e+00 - ETA: 299s - loss: 4.8904 - acc: 0.0000e+00 - ETA: 295s - loss: 4.8888 - acc: 0.0000e+00 - ETA: 296s - loss: 4.8877 - acc: 0.0000e+00 - ETA: 294s - loss: 4.8894 - acc: 0.0000e+00 - ETA: 291s - loss: 4.8895 - acc: 0.0000e+00 - ETA: 289s - loss: 4.8892 - acc: 0.0000e+00 - ETA: 287s - loss: 4.8896 - acc: 0.0000e+00 - ETA: 287s - loss: 4.8905 - acc: 0.0029     - ETA: 285s - loss: 4.8900 - acc: 0.0028 - ETA: 283s - loss: 4.8894 - acc: 0.0026 - ETA: 281s - loss: 4.8901 - acc: 0.0025 - ETA: 280s - loss: 4.8898 - acc: 0.0048 - ETA: 279s - loss: 4.8895 - acc: 0.0045 - ETA: 278s - loss: 4.8887 - acc: 0.0043 - ETA: 277s - loss: 4.8886 - acc: 0.0042 - ETA: 276s - loss: 4.8883 - acc: 0.0040 - ETA: 275s - loss: 4.8881 - acc: 0.0077 - ETA: 274s - loss: 4.8885 - acc: 0.0074 - ETA: 274s - loss: 4.8884 - acc: 0.0071 - ETA: 272s - loss: 4.8864 - acc: 0.0069 - ETA: 271s - loss: 4.8869 - acc: 0.0067 - ETA: 271s - loss: 4.8872 - acc: 0.0065 - ETA: 270s - loss: 4.8871 - acc: 0.0078 - ETA: 269s - loss: 4.8862 - acc: 0.0076 - ETA: 269s - loss: 4.8862 - acc: 0.0074 - ETA: 268s - loss: 4.8851 - acc: 0.0071 - ETA: 268s - loss: 4.8855 - acc: 0.0069 - ETA: 267s - loss: 4.8859 - acc: 0.0068 - ETA: 267s - loss: 4.8864 - acc: 0.0066 - ETA: 266s - loss: 4.8862 - acc: 0.0064 - ETA: 265s - loss: 4.8867 - acc: 0.0063 - ETA: 265s - loss: 4.8866 - acc: 0.0073 - ETA: 264s - loss: 4.8870 - acc: 0.0071 - ETA: 263s - loss: 4.8877 - acc: 0.0070 - ETA: 262s - loss: 4.8876 - acc: 0.0068 - ETA: 261s - loss: 4.8886 - acc: 0.0067 - ETA: 260s - loss: 4.8888 - acc: 0.0065 - ETA: 260s - loss: 4.8886 - acc: 0.0064 - ETA: 259s - loss: 4.8890 - acc: 0.0063 - ETA: 258s - loss: 4.8894 - acc: 0.0061 - ETA: 257s - loss: 4.8901 - acc: 0.0060 - ETA: 256s - loss: 4.8902 - acc: 0.0069 - ETA: 255s - loss: 4.8906 - acc: 0.0067 - ETA: 254s - loss: 4.8910 - acc: 0.0066 - ETA: 253s - loss: 4.8910 - acc: 0.0065 - ETA: 252s - loss: 4.8908 - acc: 0.0064 - ETA: 251s - loss: 4.8909 - acc: 0.0063 - ETA: 250s - loss: 4.8909 - acc: 0.0061 - ETA: 249s - loss: 4.8908 - acc: 0.0060 - ETA: 248s - loss: 4.8909 - acc: 0.0059 - ETA: 247s - loss: 4.8907 - acc: 0.0058 - ETA: 247s - loss: 4.8906 - acc: 0.0057 - ETA: 246s - loss: 4.8908 - acc: 0.0056 - ETA: 245s - loss: 4.8908 - acc: 0.0056 - ETA: 245s - loss: 4.8906 - acc: 0.0055 - ETA: 244s - loss: 4.8906 - acc: 0.0062 - ETA: 243s - loss: 4.8907 - acc: 0.0061 - ETA: 242s - loss: 4.8907 - acc: 0.0060 - ETA: 242s - loss: 4.8907 - acc: 0.0059 - ETA: 241s - loss: 4.8906 - acc: 0.0058 - ETA: 240s - loss: 4.8908 - acc: 0.0057 - ETA: 239s - loss: 4.8910 - acc: 0.0056 - ETA: 238s - loss: 4.8908 - acc: 0.0069 - ETA: 237s - loss: 4.8908 - acc: 0.0068 - ETA: 236s - loss: 4.8909 - acc: 0.0068 - ETA: 235s - loss: 4.8908 - acc: 0.0067 - ETA: 234s - loss: 4.8908 - acc: 0.0066 - ETA: 233s - loss: 4.8905 - acc: 0.0065 - ETA: 232s - loss: 4.8901 - acc: 0.0071 - ETA: 232s - loss: 4.8902 - acc: 0.0070 - ETA: 231s - loss: 4.8903 - acc: 0.0069 - ETA: 230s - loss: 4.8905 - acc: 0.0068 - ETA: 229s - loss: 4.8906 - acc: 0.0067 - ETA: 228s - loss: 4.8905 - acc: 0.0072 - ETA: 227s - loss: 4.8905 - acc: 0.0077 - ETA: 227s - loss: 4.8904 - acc: 0.0076 - ETA: 226s - loss: 4.8903 - acc: 0.0076 - ETA: 225s - loss: 4.8904 - acc: 0.0075 - ETA: 224s - loss: 4.8905 - acc: 0.0074 - ETA: 223s - loss: 4.8904 - acc: 0.0073 - ETA: 223s - loss: 4.8904 - acc: 0.0072 - ETA: 222s - loss: 4.8904 - acc: 0.0071 - ETA: 221s - loss: 4.8904 - acc: 0.0076 - ETA: 220s - loss: 4.8906 - acc: 0.0081 - ETA: 219s - loss: 4.8904 - acc: 0.0080 - ETA: 218s - loss: 4.8902 - acc: 0.0079 - ETA: 217s - loss: 4.8900 - acc: 0.0078 - ETA: 216s - loss: 4.8899 - acc: 0.0077 - ETA: 215s - loss: 4.8898 - acc: 0.0077 - ETA: 214s - loss: 4.8898 - acc: 0.0076 - ETA: 213s - loss: 4.8900 - acc: 0.0075 - ETA: 212s - loss: 4.8898 - acc: 0.0074 - ETA: 211s - loss: 4.8898 - acc: 0.0074 - ETA: 210s - loss: 4.8898 - acc: 0.0073 - ETA: 209s - loss: 4.8895 - acc: 0.0072 - ETA: 208s - loss: 4.8897 - acc: 0.0071 - ETA: 207s - loss: 4.8897 - acc: 0.0071 - ETA: 206s - loss: 4.8897 - acc: 0.0070 - ETA: 205s - loss: 4.8897 - acc: 0.0074 - ETA: 205s - loss: 4.8898 - acc: 0.0073 - ETA: 204s - loss: 4.8897 - acc: 0.0073 - ETA: 203s - loss: 4.8897 - acc: 0.0072 - ETA: 201s - loss: 4.8898 - acc: 0.0071 - ETA: 201s - loss: 4.8897 - acc: 0.0071 - ETA: 200s - loss: 4.8895 - acc: 0.0070 - ETA: 199s - loss: 4.8896 - acc: 0.0070 - ETA: 198s - loss: 4.8896 - acc: 0.0069 - ETA: 197s - loss: 4.8896 - acc: 0.0073 - ETA: 196s - loss: 4.8891 - acc: 0.0076 - ETA: 195s - loss: 4.8890 - acc: 0.0076 - ETA: 194s - loss: 4.8890 - acc: 0.0075 - ETA: 193s - loss: 4.8889 - acc: 0.0074 - ETA: 192s - loss: 4.8887 - acc: 0.0074 - ETA: 191s - loss: 4.8887 - acc: 0.0073 - ETA: 191s - loss: 4.8887 - acc: 0.0077 - ETA: 190s - loss: 4.8888 - acc: 0.0076 - ETA: 189s - loss: 4.8884 - acc: 0.0075 - ETA: 188s - loss: 4.8884 - acc: 0.0079 - ETA: 187s - loss: 4.8886 - acc: 0.0082 - ETA: 186s - loss: 4.8887 - acc: 0.0081 - ETA: 185s - loss: 4.8887 - acc: 0.0081 - ETA: 184s - loss: 4.8884 - acc: 0.0084 - ETA: 183s - loss: 4.8882 - acc: 0.0087 - ETA: 182s - loss: 4.8879 - acc: 0.0086 - ETA: 181s - loss: 4.8878 - acc: 0.0090 - ETA: 180s - loss: 4.8876 - acc: 0.0089 - ETA: 180s - loss: 4.8876 - acc: 0.0088 - ETA: 179s - loss: 4.8876 - acc: 0.0091 - ETA: 178s - loss: 4.8874 - acc: 0.0091 - ETA: 177s - loss: 4.8872 - acc: 0.0090 - ETA: 176s - loss: 4.8875 - acc: 0.0089 - ETA: 175s - loss: 4.8874 - acc: 0.0092 - ETA: 174s - loss: 4.8874 - acc: 0.0092 - ETA: 173s - loss: 4.8873 - acc: 0.0091 - ETA: 172s - loss: 4.8871 - acc: 0.0090 - ETA: 171s - loss: 4.8872 - acc: 0.0090 - ETA: 170s - loss: 4.8870 - acc: 0.0092 - ETA: 170s - loss: 4.8874 - acc: 0.0092 - ETA: 169s - loss: 4.8875 - acc: 0.0091 - ETA: 168s - loss: 4.8878 - acc: 0.0091 - ETA: 167s - loss: 4.8876 - acc: 0.0090 - ETA: 166s - loss: 4.8873 - acc: 0.0089 - ETA: 165s - loss: 4.8872 - acc: 0.0089 - ETA: 164s - loss: 4.8874 - acc: 0.0088 - ETA: 163s - loss: 4.8878 - acc: 0.0088 - ETA: 162s - loss: 4.8877 - acc: 0.0087 - ETA: 162s - loss: 4.8876 - acc: 0.0087 - ETA: 161s - loss: 4.8878 - acc: 0.0086 - ETA: 160s - loss: 4.8880 - acc: 0.0085 - ETA: 159s - loss: 4.8880 - acc: 0.0085 - ETA: 158s - loss: 4.8880 - acc: 0.0088 - ETA: 157s - loss: 4.8879 - acc: 0.0087 - ETA: 156s - loss: 4.8879 - acc: 0.0086 - ETA: 155s - loss: 4.8879 - acc: 0.0086 - ETA: 154s - loss: 4.8880 - acc: 0.0085 - ETA: 153s - loss: 4.8880 - acc: 0.0085 - ETA: 153s - loss: 4.8881 - acc: 0.0084 - ETA: 152s - loss: 4.8879 - acc: 0.0084 - ETA: 151s - loss: 4.8876 - acc: 0.0083 - ETA: 150s - loss: 4.8873 - acc: 0.0083 - ETA: 149s - loss: 4.8873 - acc: 0.0082 - ETA: 148s - loss: 4.8876 - acc: 0.0082 - ETA: 147s - loss: 4.8873 - acc: 0.0081 - ETA: 146s - loss: 4.8872 - acc: 0.0081 - ETA: 145s - loss: 4.8872 - acc: 0.0080 - ETA: 144s - loss: 4.8871 - acc: 0.0080 - ETA: 143s - loss: 4.8872 - acc: 0.0080 - ETA: 142s - loss: 4.8871 - acc: 0.0079 - ETA: 142s - loss: 4.8871 - acc: 0.0079 - ETA: 141s - loss: 4.8873 - acc: 0.0078 - ETA: 140s - loss: 4.8875 - acc: 0.0078 - ETA: 139s - loss: 4.8874 - acc: 0.0077 - ETA: 138s - loss: 4.8872 - acc: 0.0077 - ETA: 137s - loss: 4.8871 - acc: 0.0077 - ETA: 136s - loss: 4.8872 - acc: 0.0076 - ETA: 135s - loss: 4.8873 - acc: 0.0076 - ETA: 134s - loss: 4.8871 - acc: 0.0075 - ETA: 133s - loss: 4.8873 - acc: 0.0075 - ETA: 132s - loss: 4.8873 - acc: 0.0074 - ETA: 132s - loss: 4.8870 - acc: 0.0077 - ETA: 131s - loss: 4.8870 - acc: 0.0076 - ETA: 130s - loss: 4.8869 - acc: 0.0076 - ETA: 129s - loss: 4.8869 - acc: 0.0076 - ETA: 128s - loss: 4.8868 - acc: 0.0075 - ETA: 127s - loss: 4.8867 - acc: 0.0077 - ETA: 126s - loss: 4.8867 - acc: 0.0077 - ETA: 125s - loss: 4.8867 - acc: 0.0077 - ETA: 124s - loss: 4.8867 - acc: 0.0076 - ETA: 124s - loss: 4.8866 - acc: 0.0076 - ETA: 123s - loss: 4.8864 - acc: 0.0075 - ETA: 122s - loss: 4.8864 - acc: 0.0075 - ETA: 121s - loss: 4.8863 - acc: 0.0075 - ETA: 120s - loss: 4.8861 - acc: 0.0074 - ETA: 119s - loss: 4.8859 - acc: 0.0074 - ETA: 119s - loss: 4.8855 - acc: 0.0074 - ETA: 118s - loss: 4.8854 - acc: 0.0073 - ETA: 117s - loss: 4.8853 - acc: 0.0073 - ETA: 116s - loss: 4.8850 - acc: 0.0075 - ETA: 115s - loss: 4.8848 - acc: 0.0075 - ETA: 114s - loss: 4.8854 - acc: 0.0074 - ETA: 113s - loss: 4.8854 - acc: 0.0074 - ETA: 113s - loss: 4.8853 - acc: 0.0073 - ETA: 112s - loss: 4.8852 - acc: 0.0073 - ETA: 111s - loss: 4.8852 - acc: 0.0073 - ETA: 110s - loss: 4.8851 - acc: 0.0075 - ETA: 109s - loss: 4.8849 - acc: 0.0074 - ETA: 109s - loss: 4.8849 - acc: 0.0074 - ETA: 108s - loss: 4.8848 - acc: 0.0074 - ETA: 107s - loss: 4.8846 - acc: 0.0076 - ETA: 106s - loss: 4.8848 - acc: 0.0075 - ETA: 105s - loss: 4.8851 - acc: 0.0075 - ETA: 104s - loss: 4.8852 - acc: 0.0075 - ETA: 103s - loss: 4.8850 - acc: 0.0074 - ETA: 102s - loss: 4.8852 - acc: 0.0074 - ETA: 101s - loss: 4.8852 - acc: 0.0074 - ETA: 100s - loss: 4.8851 - acc: 0.0073 - ETA: 99s - loss: 4.8852 - acc: 0.0075  - ETA: 98s - loss: 4.8853 - acc: 0.0075 - ETA: 98s - loss: 4.8851 - acc: 0.0075 - ETA: 97s - loss: 4.8852 - acc: 0.0074 - ETA: 96s - loss: 4.8852 - acc: 0.0076 - ETA: 95s - loss: 4.8850 - acc: 0.0076 - ETA: 94s - loss: 4.8852 - acc: 0.0075 - ETA: 93s - loss: 4.8850 - acc: 0.0075 - ETA: 92s - loss: 4.8849 - acc: 0.0075 - ETA: 91s - loss: 4.8851 - acc: 0.0074 - ETA: 90s - loss: 4.8849 - acc: 0.0076 - ETA: 89s - loss: 4.8852 - acc: 0.0076 - ETA: 88s - loss: 4.8855 - acc: 0.0076 - ETA: 88s - loss: 4.8855 - acc: 0.0077 - ETA: 87s - loss: 4.8853 - acc: 0.0081 - ETA: 86s - loss: 4.8852 - acc: 0.0081 - ETA: 85s - loss: 4.8853 - acc: 0.0081 - ETA: 84s - loss: 4.8850 - acc: 0.0082 - ETA: 83s - loss: 4.8849 - acc: 0.0082 - ETA: 82s - loss: 4.8849 - acc: 0.0082 - ETA: 81s - loss: 4.8848 - acc: 0.0081 - ETA: 80s - loss: 4.8847 - acc: 0.0081 - ETA: 79s - loss: 4.8848 - acc: 0.0081 - ETA: 78s - loss: 4.8849 - acc: 0.0080 - ETA: 77s - loss: 4.8849 - acc: 0.0080 - ETA: 76s - loss: 4.8848 - acc: 0.0080 - ETA: 75s - loss: 4.8846 - acc: 0.0079 - ETA: 75s - loss: 4.8845 - acc: 0.0083 - ETA: 74s - loss: 4.8844 - acc: 0.0083 - ETA: 73s - loss: 4.8842 - acc: 0.0082 - ETA: 72s - loss: 4.8840 - acc: 0.0082 - ETA: 71s - loss: 4.8839 - acc: 0.0082 - ETA: 70s - loss: 4.8837 - acc: 0.0081 - ETA: 69s - loss: 4.8837 - acc: 0.0081 - ETA: 68s - loss: 4.8837 - acc: 0.0081 - ETA: 67s - loss: 4.8835 - acc: 0.0080 - ETA: 66s - loss: 4.8838 - acc: 0.0080 - ETA: 65s - loss: 4.8839 - acc: 0.0080 - ETA: 64s - loss: 4.8841 - acc: 0.0080 - ETA: 64s - loss: 4.8841 - acc: 0.0079 - ETA: 63s - loss: 4.8844 - acc: 0.0079 - ETA: 62s - loss: 4.8844 - acc: 0.0079 - ETA: 61s - loss: 4.8844 - acc: 0.0078 - ETA: 60s - loss: 4.8844 - acc: 0.0078 - ETA: 59s - loss: 4.8844 - acc: 0.0078 - ETA: 58s - loss: 4.8845 - acc: 0.0079 - ETA: 57s - loss: 4.8845 - acc: 0.0079 - ETA: 56s - loss: 4.8844 - acc: 0.0079 - ETA: 55s - loss: 4.8844 - acc: 0.0078 - ETA: 54s - loss: 4.8845 - acc: 0.0078 - ETA: 53s - loss: 4.8845 - acc: 0.0078 - ETA: 52s - loss: 4.8845 - acc: 0.0078 - ETA: 52s - loss: 4.8844 - acc: 0.0077 - ETA: 51s - loss: 4.8844 - acc: 0.0079 - ETA: 50s - loss: 4.8845 - acc: 0.0079 - ETA: 49s - loss: 4.8843 - acc: 0.0078 - ETA: 48s - loss: 4.8843 - acc: 0.0080 - ETA: 47s - loss: 4.8843 - acc: 0.0080 - ETA: 46s - loss: 4.8845 - acc: 0.0079 - ETA: 45s - loss: 4.8845 - acc: 0.0079 - ETA: 44s - loss: 4.8845 - acc: 0.0079 - ETA: 43s - loss: 4.8845 - acc: 0.0080 - ETA: 42s - loss: 4.8846 - acc: 0.0082 - ETA: 41s - loss: 4.8845 - acc: 0.0081 - ETA: 40s - loss: 4.8845 - acc: 0.0081 - ETA: 39s - loss: 4.8845 - acc: 0.0081 - ETA: 39s - loss: 4.8844 - acc: 0.0082 - ETA: 38s - loss: 4.8845 - acc: 0.0082 - ETA: 37s - loss: 4.8844 - acc: 0.0083 - ETA: 36s - loss: 4.8844 - acc: 0.0083 - ETA: 35s - loss: 4.8844 - acc: 0.0083 - ETA: 34s - loss: 4.8845 - acc: 0.0082 - ETA: 33s - loss: 4.8844 - acc: 0.0082 - ETA: 32s - loss: 4.8845 - acc: 0.0082 - ETA: 31s - loss: 4.8843 - acc: 0.0082 - ETA: 30s - loss: 4.8842 - acc: 0.0081 - ETA: 29s - loss: 4.8843 - acc: 0.0081 - ETA: 28s - loss: 4.8843 - acc: 0.0083 - ETA: 27s - loss: 4.8842 - acc: 0.0082 - ETA: 26s - loss: 4.8842 - acc: 0.0082 - ETA: 26s - loss: 4.8843 - acc: 0.0082 - ETA: 25s - loss: 4.8842 - acc: 0.0081 - ETA: 24s - loss: 4.8841 - acc: 0.0084 - ETA: 23s - loss: 4.8841 - acc: 0.0086 - ETA: 22s - loss: 4.8840 - acc: 0.0085 - ETA: 21s - loss: 4.8841 - acc: 0.0085 - ETA: 20s - loss: 4.8841 - acc: 0.0085 - ETA: 19s - loss: 4.8840 - acc: 0.0088 - ETA: 18s - loss: 4.8840 - acc: 0.0088 - ETA: 17s - loss: 4.8842 - acc: 0.0087 - ETA: 16s - loss: 4.8842 - acc: 0.0087 - ETA: 15s - loss: 4.8841 - acc: 0.0087 - ETA: 14s - loss: 4.8841 - acc: 0.0086 - ETA: 13s - loss: 4.8840 - acc: 0.0086 - ETA: 13s - loss: 4.8840 - acc: 0.0086 - ETA: 12s - loss: 4.8840 - acc: 0.0086 - ETA: 11s - loss: 4.8840 - acc: 0.0085 - ETA: 10s - loss: 4.8841 - acc: 0.0085 - ETA: 9s - loss: 4.8840 - acc: 0.0086  - ETA: 8s - loss: 4.8841 - acc: 0.0086 - ETA: 7s - loss: 4.8841 - acc: 0.0087 - ETA: 6s - loss: 4.8841 - acc: 0.0087 - ETA: 5s - loss: 4.8841 - acc: 0.0087 - ETA: 4s - loss: 4.8839 - acc: 0.0087 - ETA: 3s - loss: 4.8839 - acc: 0.0086 - ETA: 2s - loss: 4.8840 - acc: 0.0086 - ETA: 1s - loss: 4.8840 - acc: 0.0086 - ETA: 0s - loss: 4.8840 - acc: 0.0086Epoch 00000: val_loss improved from inf to 4.86952, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 327s - loss: 4.8839 - acc: 0.0085 - val_loss: 4.8695 - val_acc: 0.0108
    Epoch 2/10
    6660/6680 [============================>.] - ETA: 292s - loss: 4.8797 - acc: 0.0000e+00 - ETA: 296s - loss: 4.8602 - acc: 0.0250     - ETA: 299s - loss: 4.8757 - acc: 0.0167 - ETA: 300s - loss: 4.8572 - acc: 0.0250 - ETA: 299s - loss: 4.8593 - acc: 0.0200 - ETA: 299s - loss: 4.8565 - acc: 0.0167 - ETA: 298s - loss: 4.8682 - acc: 0.0143 - ETA: 297s - loss: 4.8738 - acc: 0.0188 - ETA: 296s - loss: 4.8749 - acc: 0.0167 - ETA: 295s - loss: 4.8753 - acc: 0.0150 - ETA: 293s - loss: 4.8785 - acc: 0.0136 - ETA: 290s - loss: 4.8755 - acc: 0.0125 - ETA: 289s - loss: 4.8760 - acc: 0.0154 - ETA: 288s - loss: 4.8787 - acc: 0.0143 - ETA: 288s - loss: 4.8811 - acc: 0.0167 - ETA: 287s - loss: 4.8800 - acc: 0.0188 - ETA: 286s - loss: 4.8804 - acc: 0.0176 - ETA: 286s - loss: 4.8777 - acc: 0.0194 - ETA: 285s - loss: 4.8777 - acc: 0.0184 - ETA: 284s - loss: 4.8808 - acc: 0.0175 - ETA: 283s - loss: 4.8794 - acc: 0.0167 - ETA: 282s - loss: 4.8778 - acc: 0.0159 - ETA: 281s - loss: 4.8805 - acc: 0.0152 - ETA: 281s - loss: 4.8788 - acc: 0.0146 - ETA: 279s - loss: 4.8763 - acc: 0.0140 - ETA: 278s - loss: 4.8745 - acc: 0.0154 - ETA: 277s - loss: 4.8726 - acc: 0.0148 - ETA: 276s - loss: 4.8703 - acc: 0.0143 - ETA: 275s - loss: 4.8712 - acc: 0.0155 - ETA: 274s - loss: 4.8703 - acc: 0.0150 - ETA: 273s - loss: 4.8707 - acc: 0.0145 - ETA: 272s - loss: 4.8703 - acc: 0.0156 - ETA: 271s - loss: 4.8707 - acc: 0.0152 - ETA: 270s - loss: 4.8714 - acc: 0.0147 - ETA: 270s - loss: 4.8697 - acc: 0.0143 - ETA: 268s - loss: 4.8716 - acc: 0.0139 - ETA: 267s - loss: 4.8725 - acc: 0.0135 - ETA: 266s - loss: 4.8716 - acc: 0.0145 - ETA: 265s - loss: 4.8729 - acc: 0.0141 - ETA: 264s - loss: 4.8733 - acc: 0.0138 - ETA: 263s - loss: 4.8728 - acc: 0.0134 - ETA: 262s - loss: 4.8711 - acc: 0.0131 - ETA: 262s - loss: 4.8705 - acc: 0.0128 - ETA: 260s - loss: 4.8699 - acc: 0.0136 - ETA: 260s - loss: 4.8696 - acc: 0.0133 - ETA: 259s - loss: 4.8699 - acc: 0.0130 - ETA: 258s - loss: 4.8688 - acc: 0.0138 - ETA: 257s - loss: 4.8692 - acc: 0.0146 - ETA: 256s - loss: 4.8699 - acc: 0.0153 - ETA: 255s - loss: 4.8694 - acc: 0.0150 - ETA: 254s - loss: 4.8698 - acc: 0.0147 - ETA: 253s - loss: 4.8701 - acc: 0.0144 - ETA: 252s - loss: 4.8701 - acc: 0.0142 - ETA: 251s - loss: 4.8714 - acc: 0.0139 - ETA: 250s - loss: 4.8710 - acc: 0.0136 - ETA: 249s - loss: 4.8702 - acc: 0.0134 - ETA: 248s - loss: 4.8699 - acc: 0.0132 - ETA: 247s - loss: 4.8697 - acc: 0.0129 - ETA: 246s - loss: 4.8708 - acc: 0.0127 - ETA: 245s - loss: 4.8711 - acc: 0.0125 - ETA: 244s - loss: 4.8710 - acc: 0.0123 - ETA: 243s - loss: 4.8707 - acc: 0.0121 - ETA: 242s - loss: 4.8712 - acc: 0.0119 - ETA: 242s - loss: 4.8703 - acc: 0.0117 - ETA: 241s - loss: 4.8705 - acc: 0.0115 - ETA: 240s - loss: 4.8714 - acc: 0.0114 - ETA: 239s - loss: 4.8720 - acc: 0.0119 - ETA: 238s - loss: 4.8730 - acc: 0.0118 - ETA: 237s - loss: 4.8739 - acc: 0.0116 - ETA: 236s - loss: 4.8742 - acc: 0.0114 - ETA: 235s - loss: 4.8745 - acc: 0.0113 - ETA: 234s - loss: 4.8741 - acc: 0.0111 - ETA: 234s - loss: 4.8738 - acc: 0.0110 - ETA: 233s - loss: 4.8732 - acc: 0.0108 - ETA: 232s - loss: 4.8722 - acc: 0.0113 - ETA: 231s - loss: 4.8724 - acc: 0.0112 - ETA: 230s - loss: 4.8726 - acc: 0.0110 - ETA: 229s - loss: 4.8718 - acc: 0.0109 - ETA: 228s - loss: 4.8716 - acc: 0.0108 - ETA: 227s - loss: 4.8711 - acc: 0.0113 - ETA: 226s - loss: 4.8711 - acc: 0.0111 - ETA: 226s - loss: 4.8712 - acc: 0.0110 - ETA: 225s - loss: 4.8707 - acc: 0.0108 - ETA: 224s - loss: 4.8697 - acc: 0.0107 - ETA: 223s - loss: 4.8697 - acc: 0.0106 - ETA: 222s - loss: 4.8685 - acc: 0.0110 - ETA: 221s - loss: 4.8686 - acc: 0.0109 - ETA: 220s - loss: 4.8670 - acc: 0.0114 - ETA: 219s - loss: 4.8675 - acc: 0.0112 - ETA: 218s - loss: 4.8677 - acc: 0.0117 - ETA: 217s - loss: 4.8683 - acc: 0.0115 - ETA: 217s - loss: 4.8688 - acc: 0.0114 - ETA: 216s - loss: 4.8683 - acc: 0.0113 - ETA: 215s - loss: 4.8682 - acc: 0.0117 - ETA: 214s - loss: 4.8675 - acc: 0.0116 - ETA: 213s - loss: 4.8667 - acc: 0.0125 - ETA: 212s - loss: 4.8663 - acc: 0.0124 - ETA: 211s - loss: 4.8663 - acc: 0.0122 - ETA: 210s - loss: 4.8654 - acc: 0.0121 - ETA: 209s - loss: 4.8659 - acc: 0.0120 - ETA: 209s - loss: 4.8665 - acc: 0.0119 - ETA: 208s - loss: 4.8668 - acc: 0.0118 - ETA: 207s - loss: 4.8680 - acc: 0.0117 - ETA: 206s - loss: 4.8678 - acc: 0.0120 - ETA: 205s - loss: 4.8684 - acc: 0.0119 - ETA: 204s - loss: 4.8689 - acc: 0.0118 - ETA: 203s - loss: 4.8683 - acc: 0.0117 - ETA: 202s - loss: 4.8682 - acc: 0.0116 - ETA: 201s - loss: 4.8678 - acc: 0.0119 - ETA: 201s - loss: 4.8675 - acc: 0.0118 - ETA: 200s - loss: 4.8670 - acc: 0.0122 - ETA: 199s - loss: 4.8670 - acc: 0.0121 - ETA: 198s - loss: 4.8672 - acc: 0.0119 - ETA: 197s - loss: 4.8667 - acc: 0.0118 - ETA: 196s - loss: 4.8656 - acc: 0.0122 - ETA: 195s - loss: 4.8648 - acc: 0.0125 - ETA: 194s - loss: 4.8635 - acc: 0.0124 - ETA: 193s - loss: 4.8638 - acc: 0.0131 - ETA: 192s - loss: 4.8647 - acc: 0.0130 - ETA: 192s - loss: 4.8643 - acc: 0.0129 - ETA: 191s - loss: 4.8651 - acc: 0.0128 - ETA: 190s - loss: 4.8644 - acc: 0.0127 - ETA: 189s - loss: 4.8654 - acc: 0.0126 - ETA: 188s - loss: 4.8659 - acc: 0.0125 - ETA: 187s - loss: 4.8663 - acc: 0.0124 - ETA: 186s - loss: 4.8660 - acc: 0.0127 - ETA: 185s - loss: 4.8669 - acc: 0.0126 - ETA: 184s - loss: 4.8669 - acc: 0.0129 - ETA: 184s - loss: 4.8664 - acc: 0.0128 - ETA: 183s - loss: 4.8670 - acc: 0.0127 - ETA: 182s - loss: 4.8668 - acc: 0.0130 - ETA: 181s - loss: 4.8664 - acc: 0.0133 - ETA: 180s - loss: 4.8665 - acc: 0.0132 - ETA: 179s - loss: 4.8674 - acc: 0.0131 - ETA: 178s - loss: 4.8670 - acc: 0.0130 - ETA: 177s - loss: 4.8671 - acc: 0.0129 - ETA: 176s - loss: 4.8671 - acc: 0.0128 - ETA: 175s - loss: 4.8668 - acc: 0.0134 - ETA: 175s - loss: 4.8672 - acc: 0.0133 - ETA: 174s - loss: 4.8679 - acc: 0.0136 - ETA: 173s - loss: 4.8680 - acc: 0.0135 - ETA: 172s - loss: 4.8681 - acc: 0.0134 - ETA: 171s - loss: 4.8681 - acc: 0.0133 - ETA: 170s - loss: 4.8684 - acc: 0.0132 - ETA: 169s - loss: 4.8684 - acc: 0.0131 - ETA: 168s - loss: 4.8687 - acc: 0.0130 - ETA: 167s - loss: 4.8689 - acc: 0.0129 - ETA: 167s - loss: 4.8690 - acc: 0.0132 - ETA: 166s - loss: 4.8687 - acc: 0.0134 - ETA: 165s - loss: 4.8688 - acc: 0.0137 - ETA: 164s - loss: 4.8687 - acc: 0.0142 - ETA: 163s - loss: 4.8685 - acc: 0.0145 - ETA: 162s - loss: 4.8679 - acc: 0.0147 - ETA: 161s - loss: 4.8678 - acc: 0.0146 - ETA: 160s - loss: 4.8673 - acc: 0.0145 - ETA: 159s - loss: 4.8677 - acc: 0.0144 - ETA: 158s - loss: 4.8680 - acc: 0.0143 - ETA: 158s - loss: 4.8680 - acc: 0.0142 - ETA: 157s - loss: 4.8677 - acc: 0.0142 - ETA: 156s - loss: 4.8676 - acc: 0.0141 - ETA: 155s - loss: 4.8674 - acc: 0.0143 - ETA: 154s - loss: 4.8670 - acc: 0.0142 - ETA: 153s - loss: 4.8668 - acc: 0.0141 - ETA: 152s - loss: 4.8671 - acc: 0.0140 - ETA: 151s - loss: 4.8672 - acc: 0.0139 - ETA: 150s - loss: 4.8674 - acc: 0.0139 - ETA: 150s - loss: 4.8676 - acc: 0.0138 - ETA: 149s - loss: 4.8679 - acc: 0.0140 - ETA: 148s - loss: 4.8682 - acc: 0.0139 - ETA: 147s - loss: 4.8683 - acc: 0.0138 - ETA: 146s - loss: 4.8678 - acc: 0.0140 - ETA: 145s - loss: 4.8677 - acc: 0.0140 - ETA: 144s - loss: 4.8676 - acc: 0.0139 - ETA: 143s - loss: 4.8686 - acc: 0.0138 - ETA: 142s - loss: 4.8682 - acc: 0.0137 - ETA: 142s - loss: 4.8684 - acc: 0.0139 - ETA: 141s - loss: 4.8683 - acc: 0.0141 - ETA: 140s - loss: 4.8686 - acc: 0.0140 - ETA: 139s - loss: 4.8686 - acc: 0.0142 - ETA: 138s - loss: 4.8687 - acc: 0.0142 - ETA: 137s - loss: 4.8686 - acc: 0.0141 - ETA: 136s - loss: 4.8686 - acc: 0.0140 - ETA: 135s - loss: 4.8682 - acc: 0.0139 - ETA: 134s - loss: 4.8683 - acc: 0.0139 - ETA: 134s - loss: 4.8682 - acc: 0.0143 - ETA: 133s - loss: 4.8684 - acc: 0.0142 - ETA: 132s - loss: 4.8680 - acc: 0.0144 - ETA: 131s - loss: 4.8683 - acc: 0.0144 - ETA: 130s - loss: 4.8685 - acc: 0.0143 - ETA: 129s - loss: 4.8681 - acc: 0.0142 - ETA: 128s - loss: 4.8678 - acc: 0.0141 - ETA: 127s - loss: 4.8680 - acc: 0.0141 - ETA: 126s - loss: 4.8679 - acc: 0.0140 - ETA: 125s - loss: 4.8678 - acc: 0.0139 - ETA: 125s - loss: 4.8680 - acc: 0.0138 - ETA: 124s - loss: 4.8682 - acc: 0.0138 - ETA: 123s - loss: 4.8681 - acc: 0.0137 - ETA: 122s - loss: 4.8682 - acc: 0.0139 - ETA: 121s - loss: 4.8684 - acc: 0.0138 - ETA: 120s - loss: 4.8683 - acc: 0.0140 - ETA: 119s - loss: 4.8685 - acc: 0.0139 - ETA: 118s - loss: 4.8684 - acc: 0.0141 - ETA: 117s - loss: 4.8685 - acc: 0.0143 - ETA: 117s - loss: 4.8684 - acc: 0.0142 - ETA: 116s - loss: 4.8684 - acc: 0.0141 - ETA: 115s - loss: 4.8685 - acc: 0.0141 - ETA: 114s - loss: 4.8683 - acc: 0.0140 - ETA: 113s - loss: 4.8686 - acc: 0.0139 - ETA: 112s - loss: 4.8685 - acc: 0.0139 - ETA: 111s - loss: 4.8684 - acc: 0.0138 - ETA: 110s - loss: 4.8684 - acc: 0.0137 - ETA: 109s - loss: 4.8683 - acc: 0.0137 - ETA: 108s - loss: 4.8685 - acc: 0.0136 - ETA: 108s - loss: 4.8684 - acc: 0.0136 - ETA: 107s - loss: 4.8686 - acc: 0.0135 - ETA: 106s - loss: 4.8692 - acc: 0.0134 - ETA: 105s - loss: 4.8689 - acc: 0.0134 - ETA: 104s - loss: 4.8688 - acc: 0.0135 - ETA: 103s - loss: 4.8687 - acc: 0.0135 - ETA: 102s - loss: 4.8687 - acc: 0.0136 - ETA: 101s - loss: 4.8694 - acc: 0.0136 - ETA: 100s - loss: 4.8695 - acc: 0.0135 - ETA: 100s - loss: 4.8695 - acc: 0.0137 - ETA: 99s - loss: 4.8692 - acc: 0.0136  - ETA: 98s - loss: 4.8691 - acc: 0.0136 - ETA: 97s - loss: 4.8691 - acc: 0.0137 - ETA: 96s - loss: 4.8692 - acc: 0.0139 - ETA: 95s - loss: 4.8692 - acc: 0.0138 - ETA: 94s - loss: 4.8693 - acc: 0.0138 - ETA: 93s - loss: 4.8696 - acc: 0.0137 - ETA: 92s - loss: 4.8693 - acc: 0.0136 - ETA: 91s - loss: 4.8695 - acc: 0.0136 - ETA: 91s - loss: 4.8698 - acc: 0.0135 - ETA: 90s - loss: 4.8697 - acc: 0.0139 - ETA: 89s - loss: 4.8697 - acc: 0.0140 - ETA: 88s - loss: 4.8696 - acc: 0.0142 - ETA: 87s - loss: 4.8696 - acc: 0.0141 - ETA: 86s - loss: 4.8695 - acc: 0.0141 - ETA: 85s - loss: 4.8694 - acc: 0.0140 - ETA: 84s - loss: 4.8693 - acc: 0.0142 - ETA: 83s - loss: 4.8692 - acc: 0.0143 - ETA: 82s - loss: 4.8690 - acc: 0.0143 - ETA: 82s - loss: 4.8694 - acc: 0.0142 - ETA: 81s - loss: 4.8691 - acc: 0.0145 - ETA: 80s - loss: 4.8690 - acc: 0.0145 - ETA: 79s - loss: 4.8694 - acc: 0.0144 - ETA: 78s - loss: 4.8695 - acc: 0.0144 - ETA: 77s - loss: 4.8696 - acc: 0.0145 - ETA: 76s - loss: 4.8696 - acc: 0.0145 - ETA: 75s - loss: 4.8697 - acc: 0.0144 - ETA: 74s - loss: 4.8696 - acc: 0.0143 - ETA: 73s - loss: 4.8695 - acc: 0.0143 - ETA: 73s - loss: 4.8695 - acc: 0.0142 - ETA: 72s - loss: 4.8697 - acc: 0.0142 - ETA: 71s - loss: 4.8696 - acc: 0.0143 - ETA: 70s - loss: 4.8698 - acc: 0.0143 - ETA: 69s - loss: 4.8699 - acc: 0.0142 - ETA: 68s - loss: 4.8700 - acc: 0.0141 - ETA: 67s - loss: 4.8701 - acc: 0.0141 - ETA: 66s - loss: 4.8701 - acc: 0.0140 - ETA: 65s - loss: 4.8699 - acc: 0.0140 - ETA: 64s - loss: 4.8697 - acc: 0.0141 - ETA: 64s - loss: 4.8697 - acc: 0.0143 - ETA: 63s - loss: 4.8698 - acc: 0.0144 - ETA: 62s - loss: 4.8697 - acc: 0.0145 - ETA: 61s - loss: 4.8695 - acc: 0.0145 - ETA: 60s - loss: 4.8692 - acc: 0.0146 - ETA: 59s - loss: 4.8690 - acc: 0.0146 - ETA: 58s - loss: 4.8692 - acc: 0.0145 - ETA: 57s - loss: 4.8689 - acc: 0.0146 - ETA: 56s - loss: 4.8690 - acc: 0.0146 - ETA: 55s - loss: 4.8686 - acc: 0.0145 - ETA: 54s - loss: 4.8689 - acc: 0.0145 - ETA: 54s - loss: 4.8688 - acc: 0.0144 - ETA: 53s - loss: 4.8688 - acc: 0.0144 - ETA: 52s - loss: 4.8694 - acc: 0.0143 - ETA: 51s - loss: 4.8694 - acc: 0.0144 - ETA: 50s - loss: 4.8695 - acc: 0.0144 - ETA: 49s - loss: 4.8695 - acc: 0.0143 - ETA: 48s - loss: 4.8695 - acc: 0.0143 - ETA: 47s - loss: 4.8694 - acc: 0.0142 - ETA: 46s - loss: 4.8695 - acc: 0.0142 - ETA: 46s - loss: 4.8694 - acc: 0.0141 - ETA: 45s - loss: 4.8692 - acc: 0.0141 - ETA: 44s - loss: 4.8693 - acc: 0.0140 - ETA: 43s - loss: 4.8696 - acc: 0.0142 - ETA: 42s - loss: 4.8694 - acc: 0.0141 - ETA: 41s - loss: 4.8694 - acc: 0.0142 - ETA: 40s - loss: 4.8697 - acc: 0.0142 - ETA: 39s - loss: 4.8697 - acc: 0.0141 - ETA: 38s - loss: 4.8700 - acc: 0.0141 - ETA: 37s - loss: 4.8702 - acc: 0.0140 - ETA: 37s - loss: 4.8704 - acc: 0.0140 - ETA: 36s - loss: 4.8704 - acc: 0.0141 - ETA: 35s - loss: 4.8704 - acc: 0.0141 - ETA: 34s - loss: 4.8706 - acc: 0.0140 - ETA: 33s - loss: 4.8706 - acc: 0.0140 - ETA: 32s - loss: 4.8705 - acc: 0.0139 - ETA: 31s - loss: 4.8702 - acc: 0.0140 - ETA: 30s - loss: 4.8703 - acc: 0.0140 - ETA: 29s - loss: 4.8704 - acc: 0.0140 - ETA: 28s - loss: 4.8704 - acc: 0.0139 - ETA: 28s - loss: 4.8703 - acc: 0.0139 - ETA: 27s - loss: 4.8702 - acc: 0.0140 - ETA: 26s - loss: 4.8701 - acc: 0.0141 - ETA: 25s - loss: 4.8701 - acc: 0.0141 - ETA: 24s - loss: 4.8699 - acc: 0.0140 - ETA: 23s - loss: 4.8701 - acc: 0.0140 - ETA: 22s - loss: 4.8701 - acc: 0.0139 - ETA: 21s - loss: 4.8701 - acc: 0.0140 - ETA: 20s - loss: 4.8702 - acc: 0.0140 - ETA: 19s - loss: 4.8701 - acc: 0.0139 - ETA: 18s - loss: 4.8700 - acc: 0.0142 - ETA: 18s - loss: 4.8700 - acc: 0.0142 - ETA: 17s - loss: 4.8699 - acc: 0.0141 - ETA: 16s - loss: 4.8697 - acc: 0.0141 - ETA: 15s - loss: 4.8695 - acc: 0.0140 - ETA: 14s - loss: 4.8692 - acc: 0.0142 - ETA: 13s - loss: 4.8690 - acc: 0.0143 - ETA: 12s - loss: 4.8690 - acc: 0.0142 - ETA: 11s - loss: 4.8692 - acc: 0.0142 - ETA: 10s - loss: 4.8692 - acc: 0.0141 - ETA: 9s - loss: 4.8691 - acc: 0.0141  - ETA: 9s - loss: 4.8691 - acc: 0.0142 - ETA: 8s - loss: 4.8691 - acc: 0.0142 - ETA: 7s - loss: 4.8690 - acc: 0.0141 - ETA: 6s - loss: 4.8693 - acc: 0.0142 - ETA: 5s - loss: 4.8694 - acc: 0.0142 - ETA: 4s - loss: 4.8696 - acc: 0.0141 - ETA: 3s - loss: 4.8697 - acc: 0.0141 - ETA: 2s - loss: 4.8695 - acc: 0.0144 - ETA: 1s - loss: 4.8694 - acc: 0.0143 - ETA: 0s - loss: 4.8694 - acc: 0.0143Epoch 00001: val_loss improved from 4.86952 to 4.85917, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 316s - loss: 4.8695 - acc: 0.0142 - val_loss: 4.8592 - val_acc: 0.0156
    Epoch 3/10
    6660/6680 [============================>.] - ETA: 292s - loss: 4.8630 - acc: 0.0000e+00 - ETA: 291s - loss: 4.8437 - acc: 0.0000e+00 - ETA: 293s - loss: 4.8230 - acc: 0.0167     - ETA: 291s - loss: 4.8242 - acc: 0.0125 - ETA: 290s - loss: 4.8433 - acc: 0.0100 - ETA: 290s - loss: 4.8337 - acc: 0.0083 - ETA: 290s - loss: 4.8323 - acc: 0.0071 - ETA: 289s - loss: 4.8378 - acc: 0.0063 - ETA: 287s - loss: 4.8374 - acc: 0.0056 - ETA: 287s - loss: 4.8306 - acc: 0.0050 - ETA: 286s - loss: 4.8270 - acc: 0.0091 - ETA: 285s - loss: 4.8295 - acc: 0.0083 - ETA: 284s - loss: 4.8250 - acc: 0.0077 - ETA: 282s - loss: 4.8312 - acc: 0.0071 - ETA: 282s - loss: 4.8361 - acc: 0.0067 - ETA: 281s - loss: 4.8351 - acc: 0.0094 - ETA: 281s - loss: 4.8351 - acc: 0.0088 - ETA: 280s - loss: 4.8405 - acc: 0.0083 - ETA: 279s - loss: 4.8429 - acc: 0.0079 - ETA: 278s - loss: 4.8439 - acc: 0.0075 - ETA: 278s - loss: 4.8431 - acc: 0.0071 - ETA: 277s - loss: 4.8450 - acc: 0.0068 - ETA: 276s - loss: 4.8475 - acc: 0.0087 - ETA: 276s - loss: 4.8460 - acc: 0.0083 - ETA: 275s - loss: 4.8481 - acc: 0.0080 - ETA: 274s - loss: 4.8487 - acc: 0.0077 - ETA: 274s - loss: 4.8480 - acc: 0.0093 - ETA: 273s - loss: 4.8489 - acc: 0.0089 - ETA: 272s - loss: 4.8520 - acc: 0.0103 - ETA: 272s - loss: 4.8535 - acc: 0.0100 - ETA: 271s - loss: 4.8513 - acc: 0.0097 - ETA: 270s - loss: 4.8513 - acc: 0.0109 - ETA: 270s - loss: 4.8509 - acc: 0.0106 - ETA: 269s - loss: 4.8501 - acc: 0.0118 - ETA: 269s - loss: 4.8507 - acc: 0.0114 - ETA: 268s - loss: 4.8512 - acc: 0.0111 - ETA: 267s - loss: 4.8503 - acc: 0.0108 - ETA: 266s - loss: 4.8490 - acc: 0.0105 - ETA: 266s - loss: 4.8500 - acc: 0.0103 - ETA: 265s - loss: 4.8499 - acc: 0.0100 - ETA: 264s - loss: 4.8505 - acc: 0.0098 - ETA: 263s - loss: 4.8488 - acc: 0.0095 - ETA: 262s - loss: 4.8493 - acc: 0.0093 - ETA: 261s - loss: 4.8490 - acc: 0.0091 - ETA: 260s - loss: 4.8515 - acc: 0.0089 - ETA: 259s - loss: 4.8505 - acc: 0.0087 - ETA: 259s - loss: 4.8512 - acc: 0.0085 - ETA: 258s - loss: 4.8505 - acc: 0.0094 - ETA: 257s - loss: 4.8500 - acc: 0.0092 - ETA: 256s - loss: 4.8510 - acc: 0.0090 - ETA: 255s - loss: 4.8522 - acc: 0.0088 - ETA: 255s - loss: 4.8521 - acc: 0.0087 - ETA: 254s - loss: 4.8510 - acc: 0.0085 - ETA: 253s - loss: 4.8508 - acc: 0.0083 - ETA: 252s - loss: 4.8502 - acc: 0.0082 - ETA: 251s - loss: 4.8485 - acc: 0.0089 - ETA: 251s - loss: 4.8493 - acc: 0.0088 - ETA: 250s - loss: 4.8499 - acc: 0.0086 - ETA: 249s - loss: 4.8503 - acc: 0.0085 - ETA: 248s - loss: 4.8503 - acc: 0.0083 - ETA: 247s - loss: 4.8511 - acc: 0.0090 - ETA: 246s - loss: 4.8535 - acc: 0.0097 - ETA: 245s - loss: 4.8540 - acc: 0.0095 - ETA: 244s - loss: 4.8535 - acc: 0.0094 - ETA: 243s - loss: 4.8532 - acc: 0.0092 - ETA: 242s - loss: 4.8537 - acc: 0.0098 - ETA: 241s - loss: 4.8542 - acc: 0.0097 - ETA: 240s - loss: 4.8566 - acc: 0.0096 - ETA: 239s - loss: 4.8565 - acc: 0.0094 - ETA: 238s - loss: 4.8557 - acc: 0.0093 - ETA: 237s - loss: 4.8563 - acc: 0.0099 - ETA: 236s - loss: 4.8561 - acc: 0.0104 - ETA: 235s - loss: 4.8553 - acc: 0.0103 - ETA: 234s - loss: 4.8557 - acc: 0.0101 - ETA: 234s - loss: 4.8550 - acc: 0.0107 - ETA: 233s - loss: 4.8535 - acc: 0.0112 - ETA: 232s - loss: 4.8551 - acc: 0.0110 - ETA: 231s - loss: 4.8566 - acc: 0.0109 - ETA: 230s - loss: 4.8571 - acc: 0.0108 - ETA: 229s - loss: 4.8574 - acc: 0.0106 - ETA: 228s - loss: 4.8575 - acc: 0.0105 - ETA: 227s - loss: 4.8574 - acc: 0.0104 - ETA: 226s - loss: 4.8577 - acc: 0.0102 - ETA: 225s - loss: 4.8583 - acc: 0.0107 - ETA: 224s - loss: 4.8569 - acc: 0.0112 - ETA: 223s - loss: 4.8555 - acc: 0.0110 - ETA: 222s - loss: 4.8560 - acc: 0.0109 - ETA: 221s - loss: 4.8559 - acc: 0.0108 - ETA: 220s - loss: 4.8562 - acc: 0.0107 - ETA: 220s - loss: 4.8555 - acc: 0.0111 - ETA: 219s - loss: 4.8554 - acc: 0.0110 - ETA: 218s - loss: 4.8552 - acc: 0.0109 - ETA: 217s - loss: 4.8566 - acc: 0.0113 - ETA: 216s - loss: 4.8563 - acc: 0.0112 - ETA: 215s - loss: 4.8551 - acc: 0.0116 - ETA: 214s - loss: 4.8553 - acc: 0.0125 - ETA: 213s - loss: 4.8540 - acc: 0.0129 - ETA: 212s - loss: 4.8553 - acc: 0.0128 - ETA: 211s - loss: 4.8548 - acc: 0.0131 - ETA: 210s - loss: 4.8560 - acc: 0.0130 - ETA: 210s - loss: 4.8566 - acc: 0.0129 - ETA: 209s - loss: 4.8567 - acc: 0.0127 - ETA: 208s - loss: 4.8563 - acc: 0.0126 - ETA: 207s - loss: 4.8556 - acc: 0.0130 - ETA: 206s - loss: 4.8547 - acc: 0.0129 - ETA: 205s - loss: 4.8542 - acc: 0.0132 - ETA: 204s - loss: 4.8538 - acc: 0.0131 - ETA: 203s - loss: 4.8541 - acc: 0.0130 - ETA: 202s - loss: 4.8541 - acc: 0.0133 - ETA: 202s - loss: 4.8535 - acc: 0.0132 - ETA: 201s - loss: 4.8532 - acc: 0.0131 - ETA: 200s - loss: 4.8525 - acc: 0.0129 - ETA: 199s - loss: 4.8528 - acc: 0.0128 - ETA: 198s - loss: 4.8527 - acc: 0.0127 - ETA: 197s - loss: 4.8530 - acc: 0.0126 - ETA: 196s - loss: 4.8531 - acc: 0.0129 - ETA: 195s - loss: 4.8526 - acc: 0.0128 - ETA: 194s - loss: 4.8521 - acc: 0.0127 - ETA: 193s - loss: 4.8528 - acc: 0.0126 - ETA: 192s - loss: 4.8530 - acc: 0.0125 - ETA: 191s - loss: 4.8529 - acc: 0.0124 - ETA: 190s - loss: 4.8528 - acc: 0.0123 - ETA: 190s - loss: 4.8525 - acc: 0.0122 - ETA: 189s - loss: 4.8533 - acc: 0.0125 - ETA: 188s - loss: 4.8535 - acc: 0.0124 - ETA: 187s - loss: 4.8534 - acc: 0.0127 - ETA: 186s - loss: 4.8525 - acc: 0.0130 - ETA: 185s - loss: 4.8524 - acc: 0.0129 - ETA: 184s - loss: 4.8522 - acc: 0.0136 - ETA: 183s - loss: 4.8520 - acc: 0.0138 - ETA: 182s - loss: 4.8524 - acc: 0.0141 - ETA: 181s - loss: 4.8526 - acc: 0.0140 - ETA: 181s - loss: 4.8526 - acc: 0.0139 - ETA: 180s - loss: 4.8522 - acc: 0.0142 - ETA: 179s - loss: 4.8523 - acc: 0.0141 - ETA: 178s - loss: 4.8531 - acc: 0.0140 - ETA: 177s - loss: 4.8533 - acc: 0.0139 - ETA: 176s - loss: 4.8532 - acc: 0.0138 - ETA: 175s - loss: 4.8531 - acc: 0.0137 - ETA: 174s - loss: 4.8529 - acc: 0.0136 - ETA: 173s - loss: 4.8526 - acc: 0.0135 - ETA: 172s - loss: 4.8526 - acc: 0.0134 - ETA: 172s - loss: 4.8524 - acc: 0.0136 - ETA: 171s - loss: 4.8520 - acc: 0.0135 - ETA: 170s - loss: 4.8514 - acc: 0.0134 - ETA: 169s - loss: 4.8510 - acc: 0.0137 - ETA: 168s - loss: 4.8501 - acc: 0.0136 - ETA: 167s - loss: 4.8494 - acc: 0.0139 - ETA: 166s - loss: 4.8496 - acc: 0.0138 - ETA: 165s - loss: 4.8490 - acc: 0.0137 - ETA: 164s - loss: 4.8492 - acc: 0.0136 - ETA: 163s - loss: 4.8495 - acc: 0.0135 - ETA: 162s - loss: 4.8501 - acc: 0.0134 - ETA: 161s - loss: 4.8500 - acc: 0.0133 - ETA: 161s - loss: 4.8492 - acc: 0.0132 - ETA: 160s - loss: 4.8492 - acc: 0.0131 - ETA: 159s - loss: 4.8495 - acc: 0.0131 - ETA: 158s - loss: 4.8504 - acc: 0.0130 - ETA: 157s - loss: 4.8504 - acc: 0.0129 - ETA: 156s - loss: 4.8506 - acc: 0.0134 - ETA: 155s - loss: 4.8505 - acc: 0.0134 - ETA: 154s - loss: 4.8497 - acc: 0.0133 - ETA: 153s - loss: 4.8494 - acc: 0.0132 - ETA: 152s - loss: 4.8493 - acc: 0.0137 - ETA: 152s - loss: 4.8491 - acc: 0.0136 - ETA: 151s - loss: 4.8488 - acc: 0.0136 - ETA: 150s - loss: 4.8497 - acc: 0.0135 - ETA: 149s - loss: 4.8493 - acc: 0.0137 - ETA: 148s - loss: 4.8491 - acc: 0.0139 - ETA: 147s - loss: 4.8497 - acc: 0.0138 - ETA: 146s - loss: 4.8501 - acc: 0.0137 - ETA: 145s - loss: 4.8495 - acc: 0.0140 - ETA: 144s - loss: 4.8494 - acc: 0.0139 - ETA: 143s - loss: 4.8492 - acc: 0.0141 - ETA: 142s - loss: 4.8493 - acc: 0.0140 - ETA: 142s - loss: 4.8494 - acc: 0.0139 - ETA: 141s - loss: 4.8501 - acc: 0.0138 - ETA: 140s - loss: 4.8503 - acc: 0.0138 - ETA: 139s - loss: 4.8501 - acc: 0.0142 - ETA: 138s - loss: 4.8503 - acc: 0.0142 - ETA: 137s - loss: 4.8498 - acc: 0.0141 - ETA: 136s - loss: 4.8497 - acc: 0.0143 - ETA: 135s - loss: 4.8498 - acc: 0.0142 - ETA: 134s - loss: 4.8502 - acc: 0.0141 - ETA: 133s - loss: 4.8512 - acc: 0.0141 - ETA: 133s - loss: 4.8512 - acc: 0.0140 - ETA: 132s - loss: 4.8512 - acc: 0.0139 - ETA: 131s - loss: 4.8513 - acc: 0.0144 - ETA: 130s - loss: 4.8515 - acc: 0.0143 - ETA: 129s - loss: 4.8514 - acc: 0.0142 - ETA: 128s - loss: 4.8514 - acc: 0.0144 - ETA: 127s - loss: 4.8513 - acc: 0.0143 - ETA: 126s - loss: 4.8519 - acc: 0.0142 - ETA: 125s - loss: 4.8523 - acc: 0.0142 - ETA: 125s - loss: 4.8524 - acc: 0.0141 - ETA: 124s - loss: 4.8526 - acc: 0.0140 - ETA: 123s - loss: 4.8525 - acc: 0.0140 - ETA: 122s - loss: 4.8524 - acc: 0.0139 - ETA: 121s - loss: 4.8523 - acc: 0.0138 - ETA: 120s - loss: 4.8523 - acc: 0.0140 - ETA: 119s - loss: 4.8526 - acc: 0.0139 - ETA: 118s - loss: 4.8528 - acc: 0.0139 - ETA: 117s - loss: 4.8527 - acc: 0.0138 - ETA: 116s - loss: 4.8530 - acc: 0.0137 - ETA: 116s - loss: 4.8529 - acc: 0.0137 - ETA: 115s - loss: 4.8534 - acc: 0.0136 - ETA: 114s - loss: 4.8535 - acc: 0.0135 - ETA: 113s - loss: 4.8535 - acc: 0.0135 - ETA: 112s - loss: 4.8538 - acc: 0.0134 - ETA: 111s - loss: 4.8536 - acc: 0.0136 - ETA: 110s - loss: 4.8534 - acc: 0.0142 - ETA: 109s - loss: 4.8534 - acc: 0.0142 - ETA: 108s - loss: 4.8536 - acc: 0.0141 - ETA: 107s - loss: 4.8538 - acc: 0.0143 - ETA: 107s - loss: 4.8537 - acc: 0.0142 - ETA: 106s - loss: 4.8534 - acc: 0.0141 - ETA: 105s - loss: 4.8537 - acc: 0.0141 - ETA: 104s - loss: 4.8537 - acc: 0.0140 - ETA: 103s - loss: 4.8532 - acc: 0.0139 - ETA: 102s - loss: 4.8529 - acc: 0.0141 - ETA: 101s - loss: 4.8529 - acc: 0.0140 - ETA: 100s - loss: 4.8524 - acc: 0.0140 - ETA: 99s - loss: 4.8519 - acc: 0.0139  - ETA: 98s - loss: 4.8516 - acc: 0.0141 - ETA: 98s - loss: 4.8515 - acc: 0.0142 - ETA: 97s - loss: 4.8516 - acc: 0.0142 - ETA: 96s - loss: 4.8523 - acc: 0.0141 - ETA: 95s - loss: 4.8520 - acc: 0.0143 - ETA: 94s - loss: 4.8521 - acc: 0.0142 - ETA: 93s - loss: 4.8519 - acc: 0.0141 - ETA: 92s - loss: 4.8523 - acc: 0.0141 - ETA: 91s - loss: 4.8522 - acc: 0.0140 - ETA: 90s - loss: 4.8526 - acc: 0.0139 - ETA: 89s - loss: 4.8529 - acc: 0.0139 - ETA: 89s - loss: 4.8527 - acc: 0.0143 - ETA: 88s - loss: 4.8525 - acc: 0.0142 - ETA: 87s - loss: 4.8525 - acc: 0.0141 - ETA: 86s - loss: 4.8522 - acc: 0.0141 - ETA: 85s - loss: 4.8523 - acc: 0.0140 - ETA: 84s - loss: 4.8526 - acc: 0.0142 - ETA: 83s - loss: 4.8530 - acc: 0.0141 - ETA: 82s - loss: 4.8530 - acc: 0.0140 - ETA: 81s - loss: 4.8530 - acc: 0.0140 - ETA: 80s - loss: 4.8531 - acc: 0.0139 - ETA: 80s - loss: 4.8534 - acc: 0.0141 - ETA: 79s - loss: 4.8532 - acc: 0.0140 - ETA: 78s - loss: 4.8529 - acc: 0.0140 - ETA: 77s - loss: 4.8529 - acc: 0.0139 - ETA: 76s - loss: 4.8531 - acc: 0.0141 - ETA: 75s - loss: 4.8530 - acc: 0.0144 - ETA: 74s - loss: 4.8530 - acc: 0.0143 - ETA: 73s - loss: 4.8527 - acc: 0.0143 - ETA: 72s - loss: 4.8523 - acc: 0.0142 - ETA: 71s - loss: 4.8517 - acc: 0.0146 - ETA: 70s - loss: 4.8516 - acc: 0.0145 - ETA: 70s - loss: 4.8514 - acc: 0.0145 - ETA: 69s - loss: 4.8509 - acc: 0.0144 - ETA: 68s - loss: 4.8511 - acc: 0.0143 - ETA: 67s - loss: 4.8513 - acc: 0.0143 - ETA: 66s - loss: 4.8513 - acc: 0.0142 - ETA: 65s - loss: 4.8511 - acc: 0.0146 - ETA: 64s - loss: 4.8514 - acc: 0.0147 - ETA: 63s - loss: 4.8513 - acc: 0.0146 - ETA: 62s - loss: 4.8516 - acc: 0.0146 - ETA: 61s - loss: 4.8511 - acc: 0.0147 - ETA: 61s - loss: 4.8513 - acc: 0.0148 - ETA: 60s - loss: 4.8513 - acc: 0.0152 - ETA: 59s - loss: 4.8514 - acc: 0.0151 - ETA: 58s - loss: 4.8513 - acc: 0.0151 - ETA: 57s - loss: 4.8516 - acc: 0.0150 - ETA: 56s - loss: 4.8514 - acc: 0.0149 - ETA: 55s - loss: 4.8513 - acc: 0.0149 - ETA: 54s - loss: 4.8510 - acc: 0.0150 - ETA: 53s - loss: 4.8509 - acc: 0.0150 - ETA: 52s - loss: 4.8507 - acc: 0.0149 - ETA: 52s - loss: 4.8508 - acc: 0.0149 - ETA: 51s - loss: 4.8508 - acc: 0.0148 - ETA: 50s - loss: 4.8508 - acc: 0.0147 - ETA: 49s - loss: 4.8507 - acc: 0.0147 - ETA: 48s - loss: 4.8505 - acc: 0.0146 - ETA: 47s - loss: 4.8506 - acc: 0.0146 - ETA: 46s - loss: 4.8509 - acc: 0.0145 - ETA: 45s - loss: 4.8505 - acc: 0.0145 - ETA: 44s - loss: 4.8506 - acc: 0.0144 - ETA: 44s - loss: 4.8505 - acc: 0.0144 - ETA: 43s - loss: 4.8507 - acc: 0.0143 - ETA: 42s - loss: 4.8507 - acc: 0.0143 - ETA: 41s - loss: 4.8508 - acc: 0.0144 - ETA: 40s - loss: 4.8508 - acc: 0.0145 - ETA: 39s - loss: 4.8514 - acc: 0.0145 - ETA: 38s - loss: 4.8517 - acc: 0.0144 - ETA: 37s - loss: 4.8517 - acc: 0.0146 - ETA: 36s - loss: 4.8517 - acc: 0.0145 - ETA: 35s - loss: 4.8514 - acc: 0.0145 - ETA: 34s - loss: 4.8513 - acc: 0.0147 - ETA: 34s - loss: 4.8510 - acc: 0.0147 - ETA: 33s - loss: 4.8511 - acc: 0.0146 - ETA: 32s - loss: 4.8510 - acc: 0.0146 - ETA: 31s - loss: 4.8510 - acc: 0.0145 - ETA: 30s - loss: 4.8511 - acc: 0.0145 - ETA: 29s - loss: 4.8510 - acc: 0.0145 - ETA: 28s - loss: 4.8511 - acc: 0.0144 - ETA: 27s - loss: 4.8512 - acc: 0.0144 - ETA: 26s - loss: 4.8511 - acc: 0.0143 - ETA: 26s - loss: 4.8511 - acc: 0.0143 - ETA: 25s - loss: 4.8509 - acc: 0.0142 - ETA: 24s - loss: 4.8510 - acc: 0.0142 - ETA: 23s - loss: 4.8505 - acc: 0.0143 - ETA: 22s - loss: 4.8504 - acc: 0.0142 - ETA: 21s - loss: 4.8502 - acc: 0.0142 - ETA: 20s - loss: 4.8500 - acc: 0.0143 - ETA: 19s - loss: 4.8497 - acc: 0.0143 - ETA: 18s - loss: 4.8497 - acc: 0.0142 - ETA: 17s - loss: 4.8498 - acc: 0.0142 - ETA: 17s - loss: 4.8494 - acc: 0.0141 - ETA: 16s - loss: 4.8495 - acc: 0.0141 - ETA: 15s - loss: 4.8497 - acc: 0.0140 - ETA: 14s - loss: 4.8492 - acc: 0.0140 - ETA: 13s - loss: 4.8492 - acc: 0.0139 - ETA: 12s - loss: 4.8494 - acc: 0.0139 - ETA: 11s - loss: 4.8493 - acc: 0.0139 - ETA: 10s - loss: 4.8490 - acc: 0.0138 - ETA: 9s - loss: 4.8488 - acc: 0.0141  - ETA: 8s - loss: 4.8493 - acc: 0.0140 - ETA: 8s - loss: 4.8493 - acc: 0.0143 - ETA: 7s - loss: 4.8494 - acc: 0.0143 - ETA: 6s - loss: 4.8492 - acc: 0.0142 - ETA: 5s - loss: 4.8491 - acc: 0.0142 - ETA: 4s - loss: 4.8489 - acc: 0.0141 - ETA: 3s - loss: 4.8490 - acc: 0.0141 - ETA: 2s - loss: 4.8490 - acc: 0.0140 - ETA: 1s - loss: 4.8490 - acc: 0.0143 - ETA: 0s - loss: 4.8493 - acc: 0.0144Epoch 00002: val_loss improved from 4.85917 to 4.84893, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 314s - loss: 4.8490 - acc: 0.0144 - val_loss: 4.8489 - val_acc: 0.0132
    Epoch 4/10
    6660/6680 [============================>.] - ETA: 295s - loss: 4.8374 - acc: 0.0000e+00 - ETA: 298s - loss: 4.8054 - acc: 0.0000e+00 - ETA: 298s - loss: 4.7965 - acc: 0.0167     - ETA: 297s - loss: 4.7989 - acc: 0.0125 - ETA: 295s - loss: 4.8090 - acc: 0.0100 - ETA: 295s - loss: 4.8145 - acc: 0.0083 - ETA: 294s - loss: 4.8217 - acc: 0.0071 - ETA: 293s - loss: 4.8243 - acc: 0.0063 - ETA: 292s - loss: 4.8290 - acc: 0.0056 - ETA: 292s - loss: 4.8212 - acc: 0.0050 - ETA: 291s - loss: 4.8233 - acc: 0.0045 - ETA: 290s - loss: 4.8160 - acc: 0.0083 - ETA: 289s - loss: 4.8146 - acc: 0.0077 - ETA: 288s - loss: 4.8114 - acc: 0.0071 - ETA: 286s - loss: 4.8197 - acc: 0.0067 - ETA: 285s - loss: 4.8193 - acc: 0.0063 - ETA: 285s - loss: 4.8168 - acc: 0.0088 - ETA: 284s - loss: 4.8110 - acc: 0.0083 - ETA: 283s - loss: 4.8112 - acc: 0.0079 - ETA: 282s - loss: 4.8083 - acc: 0.0100 - ETA: 281s - loss: 4.8020 - acc: 0.0167 - ETA: 280s - loss: 4.7969 - acc: 0.0182 - ETA: 280s - loss: 4.7940 - acc: 0.0174 - ETA: 279s - loss: 4.7962 - acc: 0.0167 - ETA: 278s - loss: 4.7987 - acc: 0.0160 - ETA: 277s - loss: 4.7995 - acc: 0.0154 - ETA: 276s - loss: 4.8068 - acc: 0.0148 - ETA: 275s - loss: 4.8033 - acc: 0.0179 - ETA: 274s - loss: 4.8097 - acc: 0.0172 - ETA: 273s - loss: 4.8092 - acc: 0.0167 - ETA: 272s - loss: 4.8137 - acc: 0.0161 - ETA: 272s - loss: 4.8151 - acc: 0.0156 - ETA: 271s - loss: 4.8165 - acc: 0.0152 - ETA: 270s - loss: 4.8146 - acc: 0.0147 - ETA: 269s - loss: 4.8170 - acc: 0.0143 - ETA: 268s - loss: 4.8153 - acc: 0.0139 - ETA: 267s - loss: 4.8144 - acc: 0.0135 - ETA: 266s - loss: 4.8134 - acc: 0.0132 - ETA: 265s - loss: 4.8117 - acc: 0.0141 - ETA: 265s - loss: 4.8080 - acc: 0.0138 - ETA: 264s - loss: 4.8081 - acc: 0.0134 - ETA: 263s - loss: 4.8132 - acc: 0.0131 - ETA: 262s - loss: 4.8143 - acc: 0.0128 - ETA: 261s - loss: 4.8151 - acc: 0.0136 - ETA: 261s - loss: 4.8130 - acc: 0.0133 - ETA: 262s - loss: 4.8140 - acc: 0.0130 - ETA: 261s - loss: 4.8145 - acc: 0.0128 - ETA: 260s - loss: 4.8143 - acc: 0.0125 - ETA: 259s - loss: 4.8153 - acc: 0.0122 - ETA: 258s - loss: 4.8153 - acc: 0.0120 - ETA: 257s - loss: 4.8151 - acc: 0.0127 - ETA: 256s - loss: 4.8158 - acc: 0.0125 - ETA: 255s - loss: 4.8165 - acc: 0.0123 - ETA: 254s - loss: 4.8170 - acc: 0.0120 - ETA: 253s - loss: 4.8180 - acc: 0.0118 - ETA: 253s - loss: 4.8167 - acc: 0.0116 - ETA: 252s - loss: 4.8165 - acc: 0.0123 - ETA: 251s - loss: 4.8173 - acc: 0.0121 - ETA: 250s - loss: 4.8158 - acc: 0.0127 - ETA: 249s - loss: 4.8167 - acc: 0.0125 - ETA: 248s - loss: 4.8167 - acc: 0.0131 - ETA: 247s - loss: 4.8195 - acc: 0.0129 - ETA: 246s - loss: 4.8189 - acc: 0.0127 - ETA: 245s - loss: 4.8203 - acc: 0.0133 - ETA: 244s - loss: 4.8204 - acc: 0.0131 - ETA: 243s - loss: 4.8183 - acc: 0.0129 - ETA: 242s - loss: 4.8177 - acc: 0.0127 - ETA: 241s - loss: 4.8186 - acc: 0.0125 - ETA: 240s - loss: 4.8157 - acc: 0.0130 - ETA: 239s - loss: 4.8153 - acc: 0.0129 - ETA: 238s - loss: 4.8149 - acc: 0.0134 - ETA: 237s - loss: 4.8159 - acc: 0.0132 - ETA: 236s - loss: 4.8158 - acc: 0.0130 - ETA: 235s - loss: 4.8150 - acc: 0.0128 - ETA: 234s - loss: 4.8173 - acc: 0.0127 - ETA: 233s - loss: 4.8166 - acc: 0.0132 - ETA: 232s - loss: 4.8173 - acc: 0.0130 - ETA: 231s - loss: 4.8166 - acc: 0.0135 - ETA: 230s - loss: 4.8163 - acc: 0.0133 - ETA: 229s - loss: 4.8177 - acc: 0.0131 - ETA: 228s - loss: 4.8155 - acc: 0.0142 - ETA: 228s - loss: 4.8163 - acc: 0.0140 - ETA: 227s - loss: 4.8181 - acc: 0.0139 - ETA: 226s - loss: 4.8184 - acc: 0.0137 - ETA: 225s - loss: 4.8194 - acc: 0.0135 - ETA: 224s - loss: 4.8199 - acc: 0.0134 - ETA: 223s - loss: 4.8203 - acc: 0.0132 - ETA: 222s - loss: 4.8211 - acc: 0.0131 - ETA: 221s - loss: 4.8211 - acc: 0.0129 - ETA: 220s - loss: 4.8208 - acc: 0.0128 - ETA: 219s - loss: 4.8209 - acc: 0.0132 - ETA: 218s - loss: 4.8207 - acc: 0.0130 - ETA: 217s - loss: 4.8200 - acc: 0.0129 - ETA: 216s - loss: 4.8194 - acc: 0.0128 - ETA: 215s - loss: 4.8183 - acc: 0.0126 - ETA: 214s - loss: 4.8188 - acc: 0.0125 - ETA: 214s - loss: 4.8177 - acc: 0.0129 - ETA: 213s - loss: 4.8194 - acc: 0.0128 - ETA: 212s - loss: 4.8186 - acc: 0.0126 - ETA: 211s - loss: 4.8181 - acc: 0.0125 - ETA: 210s - loss: 4.8183 - acc: 0.0124 - ETA: 209s - loss: 4.8179 - acc: 0.0123 - ETA: 208s - loss: 4.8181 - acc: 0.0126 - ETA: 207s - loss: 4.8184 - acc: 0.0125 - ETA: 206s - loss: 4.8182 - acc: 0.0124 - ETA: 205s - loss: 4.8186 - acc: 0.0123 - ETA: 204s - loss: 4.8184 - acc: 0.0121 - ETA: 204s - loss: 4.8170 - acc: 0.0120 - ETA: 203s - loss: 4.8182 - acc: 0.0119 - ETA: 202s - loss: 4.8186 - acc: 0.0118 - ETA: 201s - loss: 4.8180 - acc: 0.0122 - ETA: 200s - loss: 4.8174 - acc: 0.0121 - ETA: 199s - loss: 4.8172 - acc: 0.0124 - ETA: 198s - loss: 4.8163 - acc: 0.0123 - ETA: 197s - loss: 4.8155 - acc: 0.0122 - ETA: 196s - loss: 4.8149 - acc: 0.0121 - ETA: 195s - loss: 4.8157 - acc: 0.0120 - ETA: 195s - loss: 4.8150 - acc: 0.0127 - ETA: 194s - loss: 4.8158 - acc: 0.0126 - ETA: 193s - loss: 4.8160 - acc: 0.0125 - ETA: 192s - loss: 4.8153 - acc: 0.0128 - ETA: 191s - loss: 4.8142 - acc: 0.0131 - ETA: 190s - loss: 4.8148 - acc: 0.0130 - ETA: 189s - loss: 4.8156 - acc: 0.0129 - ETA: 188s - loss: 4.8152 - acc: 0.0128 - ETA: 187s - loss: 4.8154 - acc: 0.0131 - ETA: 187s - loss: 4.8152 - acc: 0.0130 - ETA: 186s - loss: 4.8152 - acc: 0.0133 - ETA: 185s - loss: 4.8153 - acc: 0.0136 - ETA: 184s - loss: 4.8142 - acc: 0.0146 - ETA: 183s - loss: 4.8141 - acc: 0.0145 - ETA: 182s - loss: 4.8136 - acc: 0.0152 - ETA: 181s - loss: 4.8137 - acc: 0.0150 - ETA: 180s - loss: 4.8137 - acc: 0.0153 - ETA: 179s - loss: 4.8147 - acc: 0.0152 - ETA: 179s - loss: 4.8141 - acc: 0.0158 - ETA: 178s - loss: 4.8127 - acc: 0.0161 - ETA: 177s - loss: 4.8129 - acc: 0.0159 - ETA: 176s - loss: 4.8151 - acc: 0.0158 - ETA: 175s - loss: 4.8149 - acc: 0.0157 - ETA: 174s - loss: 4.8151 - acc: 0.0156 - ETA: 173s - loss: 4.8148 - acc: 0.0158 - ETA: 172s - loss: 4.8150 - acc: 0.0161 - ETA: 171s - loss: 4.8138 - acc: 0.0163 - ETA: 170s - loss: 4.8140 - acc: 0.0162 - ETA: 169s - loss: 4.8137 - acc: 0.0171 - ETA: 168s - loss: 4.8129 - acc: 0.0170 - ETA: 167s - loss: 4.8133 - acc: 0.0169 - ETA: 166s - loss: 4.8128 - acc: 0.0171 - ETA: 166s - loss: 4.8133 - acc: 0.0173 - ETA: 165s - loss: 4.8141 - acc: 0.0172 - ETA: 164s - loss: 4.8141 - acc: 0.0174 - ETA: 163s - loss: 4.8136 - acc: 0.0173 - ETA: 162s - loss: 4.8141 - acc: 0.0172 - ETA: 161s - loss: 4.8136 - acc: 0.0174 - ETA: 160s - loss: 4.8137 - acc: 0.0176 - ETA: 159s - loss: 4.8127 - acc: 0.0178 - ETA: 158s - loss: 4.8133 - acc: 0.0177 - ETA: 157s - loss: 4.8132 - acc: 0.0176 - ETA: 156s - loss: 4.8136 - acc: 0.0175 - ETA: 155s - loss: 4.8136 - acc: 0.0174 - ETA: 155s - loss: 4.8145 - acc: 0.0173 - ETA: 154s - loss: 4.8145 - acc: 0.0172 - ETA: 153s - loss: 4.8149 - acc: 0.0171 - ETA: 152s - loss: 4.8160 - acc: 0.0170 - ETA: 151s - loss: 4.8162 - acc: 0.0169 - ETA: 150s - loss: 4.8158 - acc: 0.0171 - ETA: 149s - loss: 4.8159 - acc: 0.0170 - ETA: 148s - loss: 4.8158 - acc: 0.0169 - ETA: 147s - loss: 4.8158 - acc: 0.0168 - ETA: 146s - loss: 4.8159 - acc: 0.0167 - ETA: 145s - loss: 4.8161 - acc: 0.0166 - ETA: 145s - loss: 4.8160 - acc: 0.0168 - ETA: 144s - loss: 4.8157 - acc: 0.0167 - ETA: 143s - loss: 4.8163 - acc: 0.0166 - ETA: 142s - loss: 4.8162 - acc: 0.0165 - ETA: 141s - loss: 4.8161 - acc: 0.0164 - ETA: 140s - loss: 4.8170 - acc: 0.0163 - ETA: 139s - loss: 4.8172 - acc: 0.0162 - ETA: 138s - loss: 4.8172 - acc: 0.0161 - ETA: 137s - loss: 4.8179 - acc: 0.0160 - ETA: 136s - loss: 4.8175 - acc: 0.0159 - ETA: 135s - loss: 4.8179 - acc: 0.0158 - ETA: 135s - loss: 4.8176 - acc: 0.0160 - ETA: 134s - loss: 4.8177 - acc: 0.0159 - ETA: 133s - loss: 4.8176 - acc: 0.0159 - ETA: 132s - loss: 4.8174 - acc: 0.0158 - ETA: 131s - loss: 4.8169 - acc: 0.0162 - ETA: 130s - loss: 4.8171 - acc: 0.0161 - ETA: 129s - loss: 4.8172 - acc: 0.0161 - ETA: 128s - loss: 4.8171 - acc: 0.0160 - ETA: 127s - loss: 4.8166 - acc: 0.0159 - ETA: 126s - loss: 4.8167 - acc: 0.0158 - ETA: 126s - loss: 4.8167 - acc: 0.0157 - ETA: 125s - loss: 4.8174 - acc: 0.0156 - ETA: 124s - loss: 4.8175 - acc: 0.0156 - ETA: 123s - loss: 4.8180 - acc: 0.0155 - ETA: 122s - loss: 4.8179 - acc: 0.0157 - ETA: 121s - loss: 4.8181 - acc: 0.0158 - ETA: 120s - loss: 4.8178 - acc: 0.0158 - ETA: 119s - loss: 4.8177 - acc: 0.0157 - ETA: 118s - loss: 4.8173 - acc: 0.0158 - ETA: 117s - loss: 4.8169 - acc: 0.0160 - ETA: 117s - loss: 4.8175 - acc: 0.0159 - ETA: 116s - loss: 4.8177 - acc: 0.0161 - ETA: 115s - loss: 4.8182 - acc: 0.0160 - ETA: 114s - loss: 4.8184 - acc: 0.0162 - ETA: 113s - loss: 4.8182 - acc: 0.0161 - ETA: 112s - loss: 4.8180 - acc: 0.0163 - ETA: 111s - loss: 4.8186 - acc: 0.0162 - ETA: 110s - loss: 4.8182 - acc: 0.0161 - ETA: 109s - loss: 4.8179 - acc: 0.0163 - ETA: 108s - loss: 4.8181 - acc: 0.0164 - ETA: 107s - loss: 4.8178 - acc: 0.0168 - ETA: 107s - loss: 4.8178 - acc: 0.0167 - ETA: 106s - loss: 4.8176 - acc: 0.0167 - ETA: 105s - loss: 4.8183 - acc: 0.0166 - ETA: 104s - loss: 4.8185 - acc: 0.0165 - ETA: 103s - loss: 4.8187 - acc: 0.0164 - ETA: 102s - loss: 4.8185 - acc: 0.0164 - ETA: 101s - loss: 4.8183 - acc: 0.0165 - ETA: 100s - loss: 4.8185 - acc: 0.0167 - ETA: 99s - loss: 4.8181 - acc: 0.0166  - ETA: 98s - loss: 4.8179 - acc: 0.0167 - ETA: 98s - loss: 4.8171 - acc: 0.0169 - ETA: 97s - loss: 4.8170 - acc: 0.0170 - ETA: 96s - loss: 4.8172 - acc: 0.0170 - ETA: 95s - loss: 4.8171 - acc: 0.0169 - ETA: 94s - loss: 4.8173 - acc: 0.0168 - ETA: 93s - loss: 4.8175 - acc: 0.0167 - ETA: 92s - loss: 4.8173 - acc: 0.0169 - ETA: 91s - loss: 4.8178 - acc: 0.0168 - ETA: 90s - loss: 4.8177 - acc: 0.0167 - ETA: 89s - loss: 4.8180 - acc: 0.0167 - ETA: 88s - loss: 4.8175 - acc: 0.0168 - ETA: 88s - loss: 4.8169 - acc: 0.0174 - ETA: 87s - loss: 4.8168 - acc: 0.0175 - ETA: 86s - loss: 4.8172 - acc: 0.0174 - ETA: 85s - loss: 4.8176 - acc: 0.0174 - ETA: 84s - loss: 4.8170 - acc: 0.0173 - ETA: 83s - loss: 4.8170 - acc: 0.0174 - ETA: 82s - loss: 4.8169 - acc: 0.0174 - ETA: 81s - loss: 4.8167 - acc: 0.0173 - ETA: 80s - loss: 4.8171 - acc: 0.0174 - ETA: 79s - loss: 4.8169 - acc: 0.0173 - ETA: 79s - loss: 4.8166 - acc: 0.0173 - ETA: 78s - loss: 4.8169 - acc: 0.0172 - ETA: 77s - loss: 4.8169 - acc: 0.0171 - ETA: 76s - loss: 4.8166 - acc: 0.0175 - ETA: 75s - loss: 4.8161 - acc: 0.0174 - ETA: 74s - loss: 4.8162 - acc: 0.0175 - ETA: 73s - loss: 4.8151 - acc: 0.0181 - ETA: 72s - loss: 4.8150 - acc: 0.0180 - ETA: 71s - loss: 4.8154 - acc: 0.0179 - ETA: 70s - loss: 4.8145 - acc: 0.0178 - ETA: 70s - loss: 4.8141 - acc: 0.0178 - ETA: 69s - loss: 4.8142 - acc: 0.0177 - ETA: 68s - loss: 4.8147 - acc: 0.0176 - ETA: 67s - loss: 4.8140 - acc: 0.0176 - ETA: 66s - loss: 4.8137 - acc: 0.0175 - ETA: 65s - loss: 4.8135 - acc: 0.0174 - ETA: 64s - loss: 4.8138 - acc: 0.0176 - ETA: 63s - loss: 4.8140 - acc: 0.0175 - ETA: 62s - loss: 4.8138 - acc: 0.0176 - ETA: 61s - loss: 4.8139 - acc: 0.0177 - ETA: 61s - loss: 4.8134 - acc: 0.0179 - ETA: 60s - loss: 4.8127 - acc: 0.0178 - ETA: 59s - loss: 4.8127 - acc: 0.0177 - ETA: 58s - loss: 4.8122 - acc: 0.0177 - ETA: 57s - loss: 4.8120 - acc: 0.0178 - ETA: 56s - loss: 4.8118 - acc: 0.0179 - ETA: 55s - loss: 4.8118 - acc: 0.0178 - ETA: 54s - loss: 4.8117 - acc: 0.0178 - ETA: 53s - loss: 4.8118 - acc: 0.0181 - ETA: 52s - loss: 4.8111 - acc: 0.0180 - ETA: 52s - loss: 4.8113 - acc: 0.0179 - ETA: 51s - loss: 4.8120 - acc: 0.0179 - ETA: 50s - loss: 4.8121 - acc: 0.0180 - ETA: 49s - loss: 4.8123 - acc: 0.0181 - ETA: 48s - loss: 4.8127 - acc: 0.0180 - ETA: 47s - loss: 4.8127 - acc: 0.0180 - ETA: 46s - loss: 4.8128 - acc: 0.0179 - ETA: 45s - loss: 4.8126 - acc: 0.0178 - ETA: 44s - loss: 4.8126 - acc: 0.0180 - ETA: 44s - loss: 4.8128 - acc: 0.0179 - ETA: 43s - loss: 4.8123 - acc: 0.0178 - ETA: 42s - loss: 4.8126 - acc: 0.0178 - ETA: 41s - loss: 4.8126 - acc: 0.0177 - ETA: 40s - loss: 4.8125 - acc: 0.0176 - ETA: 39s - loss: 4.8122 - acc: 0.0178 - ETA: 38s - loss: 4.8126 - acc: 0.0177 - ETA: 37s - loss: 4.8125 - acc: 0.0178 - ETA: 36s - loss: 4.8127 - acc: 0.0177 - ETA: 35s - loss: 4.8125 - acc: 0.0177 - ETA: 35s - loss: 4.8130 - acc: 0.0176 - ETA: 34s - loss: 4.8129 - acc: 0.0176 - ETA: 33s - loss: 4.8131 - acc: 0.0177 - ETA: 32s - loss: 4.8133 - acc: 0.0178 - ETA: 31s - loss: 4.8128 - acc: 0.0179 - ETA: 30s - loss: 4.8129 - acc: 0.0178 - ETA: 29s - loss: 4.8129 - acc: 0.0178 - ETA: 28s - loss: 4.8129 - acc: 0.0177 - ETA: 27s - loss: 4.8130 - acc: 0.0177 - ETA: 26s - loss: 4.8132 - acc: 0.0176 - ETA: 26s - loss: 4.8132 - acc: 0.0175 - ETA: 25s - loss: 4.8132 - acc: 0.0175 - ETA: 24s - loss: 4.8133 - acc: 0.0176 - ETA: 23s - loss: 4.8127 - acc: 0.0177 - ETA: 22s - loss: 4.8133 - acc: 0.0176 - ETA: 21s - loss: 4.8127 - acc: 0.0177 - ETA: 20s - loss: 4.8127 - acc: 0.0177 - ETA: 19s - loss: 4.8127 - acc: 0.0176 - ETA: 18s - loss: 4.8128 - acc: 0.0176 - ETA: 17s - loss: 4.8129 - acc: 0.0175 - ETA: 17s - loss: 4.8132 - acc: 0.0175 - ETA: 16s - loss: 4.8134 - acc: 0.0174 - ETA: 15s - loss: 4.8131 - acc: 0.0174 - ETA: 14s - loss: 4.8128 - acc: 0.0173 - ETA: 13s - loss: 4.8126 - acc: 0.0174 - ETA: 12s - loss: 4.8123 - acc: 0.0173 - ETA: 11s - loss: 4.8124 - acc: 0.0173 - ETA: 10s - loss: 4.8127 - acc: 0.0172 - ETA: 9s - loss: 4.8126 - acc: 0.0173  - ETA: 8s - loss: 4.8125 - acc: 0.0174 - ETA: 8s - loss: 4.8126 - acc: 0.0175 - ETA: 7s - loss: 4.8124 - acc: 0.0175 - ETA: 6s - loss: 4.8129 - acc: 0.0174 - ETA: 5s - loss: 4.8126 - acc: 0.0175 - ETA: 4s - loss: 4.8124 - acc: 0.0175 - ETA: 3s - loss: 4.8125 - acc: 0.0174 - ETA: 2s - loss: 4.8124 - acc: 0.0175 - ETA: 1s - loss: 4.8125 - acc: 0.0175 - ETA: 0s - loss: 4.8126 - acc: 0.0174Epoch 00003: val_loss improved from 4.84893 to 4.80936, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 315s - loss: 4.8124 - acc: 0.0174 - val_loss: 4.8094 - val_acc: 0.0192
    Epoch 5/10
    6660/6680 [============================>.] - ETA: 311s - loss: 4.7606 - acc: 0.0000e+00 - ETA: 310s - loss: 4.7448 - acc: 0.0000e+00 - ETA: 306s - loss: 4.7427 - acc: 0.0000e+00 - ETA: 306s - loss: 4.8120 - acc: 0.0000e+00 - ETA: 306s - loss: 4.8071 - acc: 0.0000e+00 - ETA: 305s - loss: 4.8332 - acc: 0.0000e+00 - ETA: 303s - loss: 4.8179 - acc: 0.0000e+00 - ETA: 300s - loss: 4.8114 - acc: 0.0063     - ETA: 298s - loss: 4.8176 - acc: 0.0056 - ETA: 296s - loss: 4.8075 - acc: 0.0050 - ETA: 295s - loss: 4.8013 - acc: 0.0045 - ETA: 294s - loss: 4.7885 - acc: 0.0042 - ETA: 292s - loss: 4.8041 - acc: 0.0038 - ETA: 291s - loss: 4.8075 - acc: 0.0036 - ETA: 290s - loss: 4.8004 - acc: 0.0067 - ETA: 289s - loss: 4.8021 - acc: 0.0063 - ETA: 288s - loss: 4.8055 - acc: 0.0059 - ETA: 287s - loss: 4.8027 - acc: 0.0083 - ETA: 286s - loss: 4.8002 - acc: 0.0079 - ETA: 285s - loss: 4.8016 - acc: 0.0150 - ETA: 284s - loss: 4.7956 - acc: 0.0167 - ETA: 283s - loss: 4.7950 - acc: 0.0159 - ETA: 282s - loss: 4.7941 - acc: 0.0174 - ETA: 281s - loss: 4.7964 - acc: 0.0188 - ETA: 280s - loss: 4.7888 - acc: 0.0220 - ETA: 279s - loss: 4.7876 - acc: 0.0231 - ETA: 278s - loss: 4.7899 - acc: 0.0222 - ETA: 277s - loss: 4.7906 - acc: 0.0232 - ETA: 276s - loss: 4.7934 - acc: 0.0224 - ETA: 275s - loss: 4.7946 - acc: 0.0217 - ETA: 274s - loss: 4.7888 - acc: 0.0210 - ETA: 273s - loss: 4.7919 - acc: 0.0203 - ETA: 272s - loss: 4.7880 - acc: 0.0197 - ETA: 271s - loss: 4.7895 - acc: 0.0206 - ETA: 270s - loss: 4.7850 - acc: 0.0200 - ETA: 269s - loss: 4.7839 - acc: 0.0194 - ETA: 269s - loss: 4.7819 - acc: 0.0189 - ETA: 268s - loss: 4.7830 - acc: 0.0197 - ETA: 267s - loss: 4.7831 - acc: 0.0192 - ETA: 266s - loss: 4.7791 - acc: 0.0188 - ETA: 265s - loss: 4.7796 - acc: 0.0183 - ETA: 264s - loss: 4.7762 - acc: 0.0190 - ETA: 263s - loss: 4.7782 - acc: 0.0186 - ETA: 262s - loss: 4.7763 - acc: 0.0182 - ETA: 261s - loss: 4.7757 - acc: 0.0189 - ETA: 260s - loss: 4.7739 - acc: 0.0185 - ETA: 259s - loss: 4.7735 - acc: 0.0191 - ETA: 258s - loss: 4.7700 - acc: 0.0208 - ETA: 258s - loss: 4.7678 - acc: 0.0204 - ETA: 257s - loss: 4.7696 - acc: 0.0200 - ETA: 256s - loss: 4.7733 - acc: 0.0196 - ETA: 255s - loss: 4.7734 - acc: 0.0192 - ETA: 254s - loss: 4.7726 - acc: 0.0208 - ETA: 253s - loss: 4.7739 - acc: 0.0204 - ETA: 252s - loss: 4.7729 - acc: 0.0200 - ETA: 251s - loss: 4.7719 - acc: 0.0205 - ETA: 250s - loss: 4.7736 - acc: 0.0202 - ETA: 249s - loss: 4.7763 - acc: 0.0198 - ETA: 248s - loss: 4.7741 - acc: 0.0203 - ETA: 247s - loss: 4.7776 - acc: 0.0200 - ETA: 247s - loss: 4.7797 - acc: 0.0197 - ETA: 246s - loss: 4.7782 - acc: 0.0194 - ETA: 245s - loss: 4.7781 - acc: 0.0198 - ETA: 244s - loss: 4.7793 - acc: 0.0195 - ETA: 243s - loss: 4.7783 - acc: 0.0192 - ETA: 242s - loss: 4.7768 - acc: 0.0197 - ETA: 241s - loss: 4.7782 - acc: 0.0194 - ETA: 240s - loss: 4.7771 - acc: 0.0191 - ETA: 239s - loss: 4.7775 - acc: 0.0188 - ETA: 238s - loss: 4.7789 - acc: 0.0186 - ETA: 237s - loss: 4.7767 - acc: 0.0183 - ETA: 236s - loss: 4.7766 - acc: 0.0194 - ETA: 235s - loss: 4.7764 - acc: 0.0192 - ETA: 234s - loss: 4.7760 - acc: 0.0196 - ETA: 233s - loss: 4.7756 - acc: 0.0193 - ETA: 232s - loss: 4.7789 - acc: 0.0191 - ETA: 232s - loss: 4.7773 - acc: 0.0208 - ETA: 231s - loss: 4.7749 - acc: 0.0205 - ETA: 230s - loss: 4.7748 - acc: 0.0209 - ETA: 229s - loss: 4.7762 - acc: 0.0206 - ETA: 228s - loss: 4.7782 - acc: 0.0204 - ETA: 227s - loss: 4.7786 - acc: 0.0201 - ETA: 226s - loss: 4.7774 - acc: 0.0205 - ETA: 225s - loss: 4.7778 - acc: 0.0208 - ETA: 224s - loss: 4.7778 - acc: 0.0206 - ETA: 223s - loss: 4.7781 - acc: 0.0203 - ETA: 222s - loss: 4.7774 - acc: 0.0201 - ETA: 221s - loss: 4.7771 - acc: 0.0205 - ETA: 220s - loss: 4.7784 - acc: 0.0202 - ETA: 220s - loss: 4.7789 - acc: 0.0200 - ETA: 219s - loss: 4.7794 - acc: 0.0198 - ETA: 218s - loss: 4.7804 - acc: 0.0196 - ETA: 217s - loss: 4.7802 - acc: 0.0194 - ETA: 216s - loss: 4.7801 - acc: 0.0197 - ETA: 215s - loss: 4.7790 - acc: 0.0200 - ETA: 214s - loss: 4.7802 - acc: 0.0198 - ETA: 213s - loss: 4.7808 - acc: 0.0196 - ETA: 212s - loss: 4.7819 - acc: 0.0194 - ETA: 211s - loss: 4.7820 - acc: 0.0202 - ETA: 210s - loss: 4.7829 - acc: 0.0200 - ETA: 209s - loss: 4.7814 - acc: 0.0198 - ETA: 209s - loss: 4.7799 - acc: 0.0196 - ETA: 208s - loss: 4.7787 - acc: 0.0199 - ETA: 207s - loss: 4.7782 - acc: 0.0197 - ETA: 206s - loss: 4.7792 - acc: 0.0195 - ETA: 205s - loss: 4.7792 - acc: 0.0198 - ETA: 204s - loss: 4.7794 - acc: 0.0201 - ETA: 203s - loss: 4.7782 - acc: 0.0204 - ETA: 202s - loss: 4.7773 - acc: 0.0202 - ETA: 201s - loss: 4.7784 - acc: 0.0200 - ETA: 200s - loss: 4.7755 - acc: 0.0203 - ETA: 199s - loss: 4.7753 - acc: 0.0205 - ETA: 198s - loss: 4.7768 - acc: 0.0208 - ETA: 197s - loss: 4.7783 - acc: 0.0206 - ETA: 196s - loss: 4.7778 - acc: 0.0204 - ETA: 196s - loss: 4.7772 - acc: 0.0207 - ETA: 195s - loss: 4.7752 - acc: 0.0205 - ETA: 194s - loss: 4.7753 - acc: 0.0203 - ETA: 193s - loss: 4.7761 - acc: 0.0202 - ETA: 192s - loss: 4.7762 - acc: 0.0200 - ETA: 191s - loss: 4.7756 - acc: 0.0202 - ETA: 190s - loss: 4.7756 - acc: 0.0205 - ETA: 189s - loss: 4.7773 - acc: 0.0203 - ETA: 188s - loss: 4.7775 - acc: 0.0202 - ETA: 187s - loss: 4.7774 - acc: 0.0200 - ETA: 186s - loss: 4.7770 - acc: 0.0202 - ETA: 186s - loss: 4.7779 - acc: 0.0201 - ETA: 185s - loss: 4.7780 - acc: 0.0199 - ETA: 184s - loss: 4.7777 - acc: 0.0198 - ETA: 183s - loss: 4.7777 - acc: 0.0204 - ETA: 182s - loss: 4.7775 - acc: 0.0210 - ETA: 181s - loss: 4.7780 - acc: 0.0208 - ETA: 180s - loss: 4.7789 - acc: 0.0207 - ETA: 179s - loss: 4.7782 - acc: 0.0209 - ETA: 178s - loss: 4.7784 - acc: 0.0211 - ETA: 177s - loss: 4.7785 - acc: 0.0213 - ETA: 176s - loss: 4.7787 - acc: 0.0212 - ETA: 176s - loss: 4.7790 - acc: 0.0210 - ETA: 175s - loss: 4.7780 - acc: 0.0212 - ETA: 174s - loss: 4.7784 - acc: 0.0211 - ETA: 173s - loss: 4.7786 - acc: 0.0209 - ETA: 172s - loss: 4.7791 - acc: 0.0208 - ETA: 171s - loss: 4.7791 - acc: 0.0206 - ETA: 170s - loss: 4.7798 - acc: 0.0208 - ETA: 169s - loss: 4.7796 - acc: 0.0207 - ETA: 168s - loss: 4.7813 - acc: 0.0205 - ETA: 167s - loss: 4.7826 - acc: 0.0204 - ETA: 167s - loss: 4.7835 - acc: 0.0206 - ETA: 166s - loss: 4.7822 - acc: 0.0208 - ETA: 165s - loss: 4.7831 - acc: 0.0210 - ETA: 164s - loss: 4.7831 - acc: 0.0212 - ETA: 163s - loss: 4.7833 - acc: 0.0214 - ETA: 162s - loss: 4.7832 - acc: 0.0212 - ETA: 161s - loss: 4.7823 - acc: 0.0211 - ETA: 160s - loss: 4.7812 - acc: 0.0216 - ETA: 159s - loss: 4.7820 - acc: 0.0215 - ETA: 158s - loss: 4.7823 - acc: 0.0217 - ETA: 158s - loss: 4.7831 - acc: 0.0215 - ETA: 157s - loss: 4.7824 - acc: 0.0217 - ETA: 156s - loss: 4.7823 - acc: 0.0219 - ETA: 155s - loss: 4.7809 - acc: 0.0227 - ETA: 154s - loss: 4.7814 - acc: 0.0225 - ETA: 153s - loss: 4.7813 - acc: 0.0224 - ETA: 152s - loss: 4.7810 - acc: 0.0223 - ETA: 151s - loss: 4.7809 - acc: 0.0224 - ETA: 150s - loss: 4.7806 - acc: 0.0223 - ETA: 150s - loss: 4.7804 - acc: 0.0225 - ETA: 149s - loss: 4.7795 - acc: 0.0223 - ETA: 148s - loss: 4.7803 - acc: 0.0222 - ETA: 147s - loss: 4.7800 - acc: 0.0221 - ETA: 146s - loss: 4.7792 - acc: 0.0219 - ETA: 145s - loss: 4.7791 - acc: 0.0221 - ETA: 144s - loss: 4.7791 - acc: 0.0223 - ETA: 143s - loss: 4.7798 - acc: 0.0221 - ETA: 142s - loss: 4.7799 - acc: 0.0220 - ETA: 141s - loss: 4.7792 - acc: 0.0222 - ETA: 141s - loss: 4.7799 - acc: 0.0223 - ETA: 140s - loss: 4.7789 - acc: 0.0225 - ETA: 139s - loss: 4.7778 - acc: 0.0232 - ETA: 138s - loss: 4.7770 - acc: 0.0231 - ETA: 137s - loss: 4.7775 - acc: 0.0229 - ETA: 136s - loss: 4.7772 - acc: 0.0228 - ETA: 135s - loss: 4.7760 - acc: 0.0227 - ETA: 134s - loss: 4.7751 - acc: 0.0228 - ETA: 133s - loss: 4.7749 - acc: 0.0227 - ETA: 132s - loss: 4.7742 - acc: 0.0228 - ETA: 132s - loss: 4.7742 - acc: 0.0230 - ETA: 131s - loss: 4.7733 - acc: 0.0229 - ETA: 130s - loss: 4.7744 - acc: 0.0228 - ETA: 129s - loss: 4.7741 - acc: 0.0232 - ETA: 128s - loss: 4.7756 - acc: 0.0230 - ETA: 127s - loss: 4.7762 - acc: 0.0229 - ETA: 126s - loss: 4.7751 - acc: 0.0231 - ETA: 125s - loss: 4.7746 - acc: 0.0232 - ETA: 124s - loss: 4.7749 - acc: 0.0236 - ETA: 123s - loss: 4.7740 - acc: 0.0237 - ETA: 122s - loss: 4.7742 - acc: 0.0236 - ETA: 122s - loss: 4.7736 - acc: 0.0237 - ETA: 121s - loss: 4.7751 - acc: 0.0236 - ETA: 120s - loss: 4.7748 - acc: 0.0238 - ETA: 119s - loss: 4.7745 - acc: 0.0241 - ETA: 118s - loss: 4.7744 - acc: 0.0243 - ETA: 117s - loss: 4.7749 - acc: 0.0241 - ETA: 116s - loss: 4.7749 - acc: 0.0243 - ETA: 115s - loss: 4.7743 - acc: 0.0241 - ETA: 114s - loss: 4.7746 - acc: 0.0243 - ETA: 114s - loss: 4.7750 - acc: 0.0242 - ETA: 113s - loss: 4.7750 - acc: 0.0240 - ETA: 112s - loss: 4.7756 - acc: 0.0239 - ETA: 111s - loss: 4.7747 - acc: 0.0238 - ETA: 110s - loss: 4.7744 - acc: 0.0237 - ETA: 109s - loss: 4.7740 - acc: 0.0238 - ETA: 108s - loss: 4.7742 - acc: 0.0237 - ETA: 107s - loss: 4.7737 - acc: 0.0238 - ETA: 106s - loss: 4.7744 - acc: 0.0237 - ETA: 105s - loss: 4.7741 - acc: 0.0236 - ETA: 105s - loss: 4.7751 - acc: 0.0237 - ETA: 104s - loss: 4.7756 - acc: 0.0236 - ETA: 103s - loss: 4.7752 - acc: 0.0237 - ETA: 102s - loss: 4.7755 - acc: 0.0236 - ETA: 101s - loss: 4.7759 - acc: 0.0235 - ETA: 100s - loss: 4.7757 - acc: 0.0236 - ETA: 99s - loss: 4.7757 - acc: 0.0235  - ETA: 98s - loss: 4.7750 - acc: 0.0234 - ETA: 97s - loss: 4.7751 - acc: 0.0233 - ETA: 96s - loss: 4.7759 - acc: 0.0232 - ETA: 96s - loss: 4.7750 - acc: 0.0233 - ETA: 95s - loss: 4.7758 - acc: 0.0232 - ETA: 94s - loss: 4.7763 - acc: 0.0231 - ETA: 93s - loss: 4.7756 - acc: 0.0230 - ETA: 92s - loss: 4.7748 - acc: 0.0232 - ETA: 91s - loss: 4.7745 - acc: 0.0231 - ETA: 90s - loss: 4.7746 - acc: 0.0234 - ETA: 89s - loss: 4.7742 - acc: 0.0233 - ETA: 88s - loss: 4.7746 - acc: 0.0232 - ETA: 88s - loss: 4.7748 - acc: 0.0233 - ETA: 87s - loss: 4.7746 - acc: 0.0234 - ETA: 86s - loss: 4.7742 - acc: 0.0237 - ETA: 85s - loss: 4.7735 - acc: 0.0238 - ETA: 84s - loss: 4.7727 - acc: 0.0240 - ETA: 83s - loss: 4.7720 - acc: 0.0241 - ETA: 82s - loss: 4.7722 - acc: 0.0240 - ETA: 81s - loss: 4.7730 - acc: 0.0239 - ETA: 80s - loss: 4.7720 - acc: 0.0240 - ETA: 79s - loss: 4.7725 - acc: 0.0239 - ETA: 79s - loss: 4.7727 - acc: 0.0238 - ETA: 78s - loss: 4.7726 - acc: 0.0239 - ETA: 77s - loss: 4.7727 - acc: 0.0240 - ETA: 76s - loss: 4.7728 - acc: 0.0239 - ETA: 75s - loss: 4.7727 - acc: 0.0240 - ETA: 74s - loss: 4.7726 - acc: 0.0241 - ETA: 73s - loss: 4.7735 - acc: 0.0240 - ETA: 72s - loss: 4.7736 - acc: 0.0239 - ETA: 71s - loss: 4.7732 - acc: 0.0240 - ETA: 70s - loss: 4.7732 - acc: 0.0239 - ETA: 70s - loss: 4.7732 - acc: 0.0238 - ETA: 69s - loss: 4.7732 - acc: 0.0237 - ETA: 68s - loss: 4.7734 - acc: 0.0236 - ETA: 67s - loss: 4.7728 - acc: 0.0237 - ETA: 66s - loss: 4.7731 - acc: 0.0237 - ETA: 65s - loss: 4.7729 - acc: 0.0236 - ETA: 64s - loss: 4.7728 - acc: 0.0237 - ETA: 63s - loss: 4.7730 - acc: 0.0236 - ETA: 62s - loss: 4.7734 - acc: 0.0235 - ETA: 62s - loss: 4.7738 - acc: 0.0234 - ETA: 61s - loss: 4.7736 - acc: 0.0235 - ETA: 60s - loss: 4.7734 - acc: 0.0236 - ETA: 59s - loss: 4.7736 - acc: 0.0235 - ETA: 58s - loss: 4.7739 - acc: 0.0234 - ETA: 57s - loss: 4.7737 - acc: 0.0233 - ETA: 56s - loss: 4.7740 - acc: 0.0232 - ETA: 55s - loss: 4.7736 - acc: 0.0235 - ETA: 54s - loss: 4.7743 - acc: 0.0234 - ETA: 53s - loss: 4.7749 - acc: 0.0237 - ETA: 53s - loss: 4.7754 - acc: 0.0236 - ETA: 52s - loss: 4.7750 - acc: 0.0236 - ETA: 51s - loss: 4.7750 - acc: 0.0235 - ETA: 50s - loss: 4.7743 - acc: 0.0237 - ETA: 49s - loss: 4.7748 - acc: 0.0238 - ETA: 48s - loss: 4.7751 - acc: 0.0238 - ETA: 47s - loss: 4.7753 - acc: 0.0237 - ETA: 46s - loss: 4.7756 - acc: 0.0238 - ETA: 45s - loss: 4.7765 - acc: 0.0237 - ETA: 44s - loss: 4.7759 - acc: 0.0238 - ETA: 44s - loss: 4.7757 - acc: 0.0237 - ETA: 43s - loss: 4.7757 - acc: 0.0236 - ETA: 42s - loss: 4.7758 - acc: 0.0237 - ETA: 41s - loss: 4.7757 - acc: 0.0236 - ETA: 40s - loss: 4.7754 - acc: 0.0237 - ETA: 39s - loss: 4.7756 - acc: 0.0236 - ETA: 38s - loss: 4.7750 - acc: 0.0237 - ETA: 37s - loss: 4.7744 - acc: 0.0238 - ETA: 36s - loss: 4.7738 - acc: 0.0237 - ETA: 36s - loss: 4.7739 - acc: 0.0236 - ETA: 35s - loss: 4.7737 - acc: 0.0237 - ETA: 34s - loss: 4.7738 - acc: 0.0236 - ETA: 33s - loss: 4.7738 - acc: 0.0237 - ETA: 32s - loss: 4.7741 - acc: 0.0238 - ETA: 31s - loss: 4.7739 - acc: 0.0239 - ETA: 30s - loss: 4.7741 - acc: 0.0238 - ETA: 29s - loss: 4.7740 - acc: 0.0238 - ETA: 28s - loss: 4.7741 - acc: 0.0237 - ETA: 27s - loss: 4.7745 - acc: 0.0236 - ETA: 27s - loss: 4.7742 - acc: 0.0235 - ETA: 26s - loss: 4.7737 - acc: 0.0234 - ETA: 25s - loss: 4.7733 - acc: 0.0234 - ETA: 24s - loss: 4.7730 - acc: 0.0233 - ETA: 23s - loss: 4.7733 - acc: 0.0232 - ETA: 22s - loss: 4.7738 - acc: 0.0231 - ETA: 21s - loss: 4.7739 - acc: 0.0231 - ETA: 20s - loss: 4.7736 - acc: 0.0233 - ETA: 19s - loss: 4.7738 - acc: 0.0232 - ETA: 18s - loss: 4.7737 - acc: 0.0235 - ETA: 18s - loss: 4.7734 - acc: 0.0237 - ETA: 17s - loss: 4.7732 - acc: 0.0238 - ETA: 16s - loss: 4.7731 - acc: 0.0237 - ETA: 15s - loss: 4.7726 - acc: 0.0237 - ETA: 14s - loss: 4.7730 - acc: 0.0236 - ETA: 13s - loss: 4.7732 - acc: 0.0235 - ETA: 12s - loss: 4.7735 - acc: 0.0234 - ETA: 11s - loss: 4.7735 - acc: 0.0235 - ETA: 10s - loss: 4.7735 - acc: 0.0236 - ETA: 9s - loss: 4.7736 - acc: 0.0235  - ETA: 9s - loss: 4.7737 - acc: 0.0235 - ETA: 8s - loss: 4.7738 - acc: 0.0235 - ETA: 7s - loss: 4.7740 - acc: 0.0235 - ETA: 6s - loss: 4.7741 - acc: 0.0234 - ETA: 5s - loss: 4.7742 - acc: 0.0236 - ETA: 4s - loss: 4.7736 - acc: 0.0237 - ETA: 3s - loss: 4.7737 - acc: 0.0236 - ETA: 2s - loss: 4.7740 - acc: 0.0236 - ETA: 1s - loss: 4.7736 - acc: 0.0236 - ETA: 0s - loss: 4.7741 - acc: 0.0236Epoch 00004: val_loss improved from 4.80936 to 4.79618, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 316s - loss: 4.7744 - acc: 0.0235 - val_loss: 4.7962 - val_acc: 0.0192
    Epoch 6/10
    6660/6680 [============================>.] - ETA: 297s - loss: 4.8165 - acc: 0.0000e+00 - ETA: 298s - loss: 4.7651 - acc: 0.0000e+00 - ETA: 301s - loss: 4.8076 - acc: 0.0000e+00 - ETA: 303s - loss: 4.8425 - acc: 0.0000e+00 - ETA: 302s - loss: 4.8252 - acc: 0.0000e+00 - ETA: 302s - loss: 4.8203 - acc: 0.0000e+00 - ETA: 301s - loss: 4.8020 - acc: 0.0071     - ETA: 302s - loss: 4.8333 - acc: 0.0063 - ETA: 300s - loss: 4.8219 - acc: 0.0056 - ETA: 299s - loss: 4.8254 - acc: 0.0100 - ETA: 297s - loss: 4.8249 - acc: 0.0091 - ETA: 297s - loss: 4.8197 - acc: 0.0125 - ETA: 296s - loss: 4.8228 - acc: 0.0154 - ETA: 294s - loss: 4.8221 - acc: 0.0143 - ETA: 293s - loss: 4.8157 - acc: 0.0200 - ETA: 291s - loss: 4.8083 - acc: 0.0219 - ETA: 291s - loss: 4.7966 - acc: 0.0206 - ETA: 290s - loss: 4.7994 - acc: 0.0194 - ETA: 289s - loss: 4.8053 - acc: 0.0237 - ETA: 288s - loss: 4.8007 - acc: 0.0225 - ETA: 287s - loss: 4.7855 - acc: 0.0214 - ETA: 286s - loss: 4.7920 - acc: 0.0205 - ETA: 286s - loss: 4.7938 - acc: 0.0196 - ETA: 285s - loss: 4.7925 - acc: 0.0208 - ETA: 284s - loss: 4.7871 - acc: 0.0220 - ETA: 283s - loss: 4.7821 - acc: 0.0212 - ETA: 282s - loss: 4.7807 - acc: 0.0204 - ETA: 281s - loss: 4.7859 - acc: 0.0196 - ETA: 280s - loss: 4.7898 - acc: 0.0190 - ETA: 279s - loss: 4.7843 - acc: 0.0200 - ETA: 278s - loss: 4.7812 - acc: 0.0210 - ETA: 277s - loss: 4.7731 - acc: 0.0219 - ETA: 276s - loss: 4.7693 - acc: 0.0227 - ETA: 275s - loss: 4.7655 - acc: 0.0250 - ETA: 274s - loss: 4.7658 - acc: 0.0243 - ETA: 274s - loss: 4.7648 - acc: 0.0236 - ETA: 273s - loss: 4.7613 - acc: 0.0230 - ETA: 272s - loss: 4.7558 - acc: 0.0224 - ETA: 271s - loss: 4.7569 - acc: 0.0218 - ETA: 270s - loss: 4.7454 - acc: 0.0238 - ETA: 269s - loss: 4.7488 - acc: 0.0244 - ETA: 268s - loss: 4.7475 - acc: 0.0250 - ETA: 267s - loss: 4.7505 - acc: 0.0256 - ETA: 266s - loss: 4.7506 - acc: 0.0250 - ETA: 265s - loss: 4.7458 - acc: 0.0256 - ETA: 264s - loss: 4.7454 - acc: 0.0250 - ETA: 263s - loss: 4.7424 - acc: 0.0255 - ETA: 262s - loss: 4.7418 - acc: 0.0260 - ETA: 261s - loss: 4.7420 - acc: 0.0255 - ETA: 260s - loss: 4.7371 - acc: 0.0250 - ETA: 259s - loss: 4.7327 - acc: 0.0255 - ETA: 258s - loss: 4.7362 - acc: 0.0250 - ETA: 258s - loss: 4.7348 - acc: 0.0245 - ETA: 256s - loss: 4.7355 - acc: 0.0241 - ETA: 256s - loss: 4.7375 - acc: 0.0236 - ETA: 255s - loss: 4.7426 - acc: 0.0241 - ETA: 254s - loss: 4.7464 - acc: 0.0237 - ETA: 253s - loss: 4.7486 - acc: 0.0233 - ETA: 252s - loss: 4.7469 - acc: 0.0229 - ETA: 251s - loss: 4.7461 - acc: 0.0225 - ETA: 250s - loss: 4.7487 - acc: 0.0221 - ETA: 249s - loss: 4.7485 - acc: 0.0218 - ETA: 248s - loss: 4.7463 - acc: 0.0214 - ETA: 247s - loss: 4.7479 - acc: 0.0211 - ETA: 246s - loss: 4.7463 - acc: 0.0208 - ETA: 245s - loss: 4.7431 - acc: 0.0220 - ETA: 244s - loss: 4.7425 - acc: 0.0224 - ETA: 243s - loss: 4.7397 - acc: 0.0221 - ETA: 242s - loss: 4.7377 - acc: 0.0225 - ETA: 242s - loss: 4.7389 - acc: 0.0221 - ETA: 241s - loss: 4.7374 - acc: 0.0225 - ETA: 240s - loss: 4.7366 - acc: 0.0229 - ETA: 239s - loss: 4.7374 - acc: 0.0226 - ETA: 238s - loss: 4.7352 - acc: 0.0236 - ETA: 237s - loss: 4.7330 - acc: 0.0240 - ETA: 236s - loss: 4.7323 - acc: 0.0243 - ETA: 235s - loss: 4.7325 - acc: 0.0247 - ETA: 234s - loss: 4.7358 - acc: 0.0250 - ETA: 233s - loss: 4.7341 - acc: 0.0259 - ETA: 232s - loss: 4.7329 - acc: 0.0256 - ETA: 231s - loss: 4.7314 - acc: 0.0253 - ETA: 230s - loss: 4.7329 - acc: 0.0250 - ETA: 229s - loss: 4.7355 - acc: 0.0253 - ETA: 229s - loss: 4.7352 - acc: 0.0250 - ETA: 228s - loss: 4.7345 - acc: 0.0247 - ETA: 227s - loss: 4.7375 - acc: 0.0244 - ETA: 226s - loss: 4.7385 - acc: 0.0241 - ETA: 225s - loss: 4.7375 - acc: 0.0239 - ETA: 224s - loss: 4.7370 - acc: 0.0242 - ETA: 223s - loss: 4.7370 - acc: 0.0239 - ETA: 222s - loss: 4.7374 - acc: 0.0242 - ETA: 221s - loss: 4.7400 - acc: 0.0239 - ETA: 220s - loss: 4.7410 - acc: 0.0242 - ETA: 219s - loss: 4.7395 - acc: 0.0239 - ETA: 219s - loss: 4.7397 - acc: 0.0242 - ETA: 218s - loss: 4.7390 - acc: 0.0240 - ETA: 217s - loss: 4.7370 - acc: 0.0242 - ETA: 216s - loss: 4.7371 - acc: 0.0240 - ETA: 215s - loss: 4.7373 - acc: 0.0237 - ETA: 214s - loss: 4.7371 - acc: 0.0235 - ETA: 213s - loss: 4.7370 - acc: 0.0238 - ETA: 212s - loss: 4.7370 - acc: 0.0235 - ETA: 211s - loss: 4.7381 - acc: 0.0238 - ETA: 210s - loss: 4.7349 - acc: 0.0245 - ETA: 210s - loss: 4.7339 - acc: 0.0248 - ETA: 209s - loss: 4.7347 - acc: 0.0245 - ETA: 208s - loss: 4.7338 - acc: 0.0243 - ETA: 207s - loss: 4.7330 - acc: 0.0245 - ETA: 206s - loss: 4.7338 - acc: 0.0243 - ETA: 205s - loss: 4.7335 - acc: 0.0245 - ETA: 204s - loss: 4.7356 - acc: 0.0248 - ETA: 203s - loss: 4.7355 - acc: 0.0246 - ETA: 202s - loss: 4.7364 - acc: 0.0252 - ETA: 201s - loss: 4.7364 - acc: 0.0254 - ETA: 200s - loss: 4.7355 - acc: 0.0257 - ETA: 199s - loss: 4.7365 - acc: 0.0259 - ETA: 199s - loss: 4.7348 - acc: 0.0256 - ETA: 197s - loss: 4.7354 - acc: 0.0254 - ETA: 197s - loss: 4.7350 - acc: 0.0252 - ETA: 196s - loss: 4.7358 - acc: 0.0254 - ETA: 195s - loss: 4.7333 - acc: 0.0252 - ETA: 194s - loss: 4.7337 - acc: 0.0250 - ETA: 193s - loss: 4.7334 - acc: 0.0252 - ETA: 192s - loss: 4.7332 - acc: 0.0250 - ETA: 191s - loss: 4.7341 - acc: 0.0248 - ETA: 190s - loss: 4.7332 - acc: 0.0246 - ETA: 189s - loss: 4.7338 - acc: 0.0244 - ETA: 188s - loss: 4.7335 - acc: 0.0246 - ETA: 187s - loss: 4.7341 - acc: 0.0244 - ETA: 187s - loss: 4.7359 - acc: 0.0242 - ETA: 186s - loss: 4.7358 - acc: 0.0244 - ETA: 185s - loss: 4.7360 - acc: 0.0242 - ETA: 184s - loss: 4.7376 - acc: 0.0241 - ETA: 183s - loss: 4.7382 - acc: 0.0239 - ETA: 182s - loss: 4.7404 - acc: 0.0237 - ETA: 181s - loss: 4.7400 - acc: 0.0239 - ETA: 180s - loss: 4.7390 - acc: 0.0241 - ETA: 179s - loss: 4.7385 - acc: 0.0239 - ETA: 178s - loss: 4.7395 - acc: 0.0237 - ETA: 177s - loss: 4.7387 - acc: 0.0239 - ETA: 176s - loss: 4.7403 - acc: 0.0238 - ETA: 176s - loss: 4.7404 - acc: 0.0236 - ETA: 175s - loss: 4.7405 - acc: 0.0234 - ETA: 174s - loss: 4.7404 - acc: 0.0233 - ETA: 173s - loss: 4.7394 - acc: 0.0234 - ETA: 172s - loss: 4.7393 - acc: 0.0233 - ETA: 171s - loss: 4.7407 - acc: 0.0231 - ETA: 170s - loss: 4.7400 - acc: 0.0230 - ETA: 169s - loss: 4.7412 - acc: 0.0228 - ETA: 168s - loss: 4.7403 - acc: 0.0227 - ETA: 167s - loss: 4.7404 - acc: 0.0228 - ETA: 166s - loss: 4.7403 - acc: 0.0227 - ETA: 165s - loss: 4.7417 - acc: 0.0225 - ETA: 165s - loss: 4.7409 - acc: 0.0231 - ETA: 164s - loss: 4.7413 - acc: 0.0229 - ETA: 163s - loss: 4.7408 - acc: 0.0231 - ETA: 162s - loss: 4.7402 - acc: 0.0232 - ETA: 161s - loss: 4.7397 - acc: 0.0234 - ETA: 160s - loss: 4.7391 - acc: 0.0233 - ETA: 159s - loss: 4.7388 - acc: 0.0231 - ETA: 158s - loss: 4.7392 - acc: 0.0230 - ETA: 157s - loss: 4.7396 - acc: 0.0228 - ETA: 156s - loss: 4.7398 - acc: 0.0227 - ETA: 155s - loss: 4.7403 - acc: 0.0226 - ETA: 155s - loss: 4.7401 - acc: 0.0224 - ETA: 154s - loss: 4.7399 - acc: 0.0223 - ETA: 153s - loss: 4.7405 - acc: 0.0222 - ETA: 152s - loss: 4.7410 - acc: 0.0220 - ETA: 151s - loss: 4.7419 - acc: 0.0219 - ETA: 150s - loss: 4.7424 - acc: 0.0218 - ETA: 149s - loss: 4.7419 - acc: 0.0225 - ETA: 148s - loss: 4.7438 - acc: 0.0227 - ETA: 147s - loss: 4.7434 - acc: 0.0234 - ETA: 146s - loss: 4.7433 - acc: 0.0233 - ETA: 145s - loss: 4.7432 - acc: 0.0231 - ETA: 144s - loss: 4.7432 - acc: 0.0230 - ETA: 143s - loss: 4.7428 - acc: 0.0229 - ETA: 142s - loss: 4.7424 - acc: 0.0230 - ETA: 142s - loss: 4.7419 - acc: 0.0229 - ETA: 141s - loss: 4.7413 - acc: 0.0228 - ETA: 140s - loss: 4.7412 - acc: 0.0229 - ETA: 139s - loss: 4.7418 - acc: 0.0228 - ETA: 138s - loss: 4.7413 - acc: 0.0230 - ETA: 137s - loss: 4.7407 - acc: 0.0228 - ETA: 136s - loss: 4.7395 - acc: 0.0230 - ETA: 135s - loss: 4.7387 - acc: 0.0228 - ETA: 134s - loss: 4.7387 - acc: 0.0233 - ETA: 133s - loss: 4.7389 - acc: 0.0231 - ETA: 132s - loss: 4.7398 - acc: 0.0230 - ETA: 131s - loss: 4.7396 - acc: 0.0232 - ETA: 130s - loss: 4.7397 - acc: 0.0230 - ETA: 130s - loss: 4.7402 - acc: 0.0229 - ETA: 129s - loss: 4.7404 - acc: 0.0231 - ETA: 128s - loss: 4.7403 - acc: 0.0235 - ETA: 127s - loss: 4.7409 - acc: 0.0233 - ETA: 126s - loss: 4.7409 - acc: 0.0235 - ETA: 125s - loss: 4.7409 - acc: 0.0234 - ETA: 124s - loss: 4.7410 - acc: 0.0232 - ETA: 123s - loss: 4.7410 - acc: 0.0234 - ETA: 122s - loss: 4.7411 - acc: 0.0233 - ETA: 121s - loss: 4.7412 - acc: 0.0231 - ETA: 120s - loss: 4.7420 - acc: 0.0230 - ETA: 119s - loss: 4.7420 - acc: 0.0229 - ETA: 118s - loss: 4.7418 - acc: 0.0230 - ETA: 118s - loss: 4.7412 - acc: 0.0229 - ETA: 117s - loss: 4.7414 - acc: 0.0228 - ETA: 116s - loss: 4.7408 - acc: 0.0229 - ETA: 115s - loss: 4.7406 - acc: 0.0228 - ETA: 114s - loss: 4.7397 - acc: 0.0227 - ETA: 113s - loss: 4.7397 - acc: 0.0226 - ETA: 112s - loss: 4.7397 - acc: 0.0225 - ETA: 111s - loss: 4.7398 - acc: 0.0224 - ETA: 110s - loss: 4.7398 - acc: 0.0223 - ETA: 109s - loss: 4.7402 - acc: 0.0224 - ETA: 108s - loss: 4.7394 - acc: 0.0226 - ETA: 107s - loss: 4.7403 - acc: 0.0225 - ETA: 107s - loss: 4.7404 - acc: 0.0224 - ETA: 106s - loss: 4.7401 - acc: 0.0222 - ETA: 105s - loss: 4.7405 - acc: 0.0224 - ETA: 104s - loss: 4.7395 - acc: 0.0223 - ETA: 103s - loss: 4.7396 - acc: 0.0222 - ETA: 102s - loss: 4.7393 - acc: 0.0221 - ETA: 101s - loss: 4.7391 - acc: 0.0222 - ETA: 100s - loss: 4.7397 - acc: 0.0221 - ETA: 99s - loss: 4.7398 - acc: 0.0220  - ETA: 98s - loss: 4.7399 - acc: 0.0219 - ETA: 97s - loss: 4.7399 - acc: 0.0218 - ETA: 96s - loss: 4.7398 - acc: 0.0217 - ETA: 96s - loss: 4.7384 - acc: 0.0221 - ETA: 95s - loss: 4.7381 - acc: 0.0220 - ETA: 94s - loss: 4.7385 - acc: 0.0221 - ETA: 93s - loss: 4.7387 - acc: 0.0220 - ETA: 92s - loss: 4.7390 - acc: 0.0219 - ETA: 91s - loss: 4.7384 - acc: 0.0218 - ETA: 90s - loss: 4.7375 - acc: 0.0219 - ETA: 89s - loss: 4.7369 - acc: 0.0218 - ETA: 88s - loss: 4.7365 - acc: 0.0217 - ETA: 87s - loss: 4.7361 - acc: 0.0216 - ETA: 86s - loss: 4.7359 - acc: 0.0215 - ETA: 86s - loss: 4.7365 - acc: 0.0215 - ETA: 85s - loss: 4.7362 - acc: 0.0214 - ETA: 84s - loss: 4.7361 - acc: 0.0215 - ETA: 83s - loss: 4.7358 - acc: 0.0216 - ETA: 82s - loss: 4.7353 - acc: 0.0215 - ETA: 81s - loss: 4.7361 - acc: 0.0214 - ETA: 80s - loss: 4.7358 - acc: 0.0215 - ETA: 79s - loss: 4.7358 - acc: 0.0215 - ETA: 78s - loss: 4.7360 - acc: 0.0214 - ETA: 77s - loss: 4.7359 - acc: 0.0215 - ETA: 76s - loss: 4.7360 - acc: 0.0214 - ETA: 76s - loss: 4.7362 - acc: 0.0213 - ETA: 75s - loss: 4.7362 - acc: 0.0214 - ETA: 74s - loss: 4.7363 - acc: 0.0217 - ETA: 73s - loss: 4.7360 - acc: 0.0219 - ETA: 72s - loss: 4.7359 - acc: 0.0220 - ETA: 71s - loss: 4.7357 - acc: 0.0219 - ETA: 70s - loss: 4.7359 - acc: 0.0218 - ETA: 69s - loss: 4.7353 - acc: 0.0219 - ETA: 68s - loss: 4.7353 - acc: 0.0218 - ETA: 67s - loss: 4.7362 - acc: 0.0219 - ETA: 66s - loss: 4.7355 - acc: 0.0218 - ETA: 65s - loss: 4.7355 - acc: 0.0219 - ETA: 65s - loss: 4.7359 - acc: 0.0219 - ETA: 64s - loss: 4.7368 - acc: 0.0220 - ETA: 63s - loss: 4.7369 - acc: 0.0219 - ETA: 62s - loss: 4.7369 - acc: 0.0218 - ETA: 61s - loss: 4.7368 - acc: 0.0219 - ETA: 60s - loss: 4.7373 - acc: 0.0220 - ETA: 59s - loss: 4.7376 - acc: 0.0219 - ETA: 58s - loss: 4.7378 - acc: 0.0219 - ETA: 57s - loss: 4.7376 - acc: 0.0218 - ETA: 56s - loss: 4.7382 - acc: 0.0219 - ETA: 55s - loss: 4.7378 - acc: 0.0218 - ETA: 54s - loss: 4.7386 - acc: 0.0217 - ETA: 54s - loss: 4.7379 - acc: 0.0218 - ETA: 53s - loss: 4.7385 - acc: 0.0217 - ETA: 52s - loss: 4.7385 - acc: 0.0217 - ETA: 51s - loss: 4.7383 - acc: 0.0218 - ETA: 50s - loss: 4.7387 - acc: 0.0217 - ETA: 49s - loss: 4.7388 - acc: 0.0216 - ETA: 48s - loss: 4.7388 - acc: 0.0215 - ETA: 47s - loss: 4.7391 - acc: 0.0216 - ETA: 46s - loss: 4.7395 - acc: 0.0217 - ETA: 45s - loss: 4.7397 - acc: 0.0217 - ETA: 44s - loss: 4.7392 - acc: 0.0219 - ETA: 43s - loss: 4.7398 - acc: 0.0219 - ETA: 43s - loss: 4.7405 - acc: 0.0218 - ETA: 42s - loss: 4.7407 - acc: 0.0217 - ETA: 41s - loss: 4.7408 - acc: 0.0216 - ETA: 40s - loss: 4.7408 - acc: 0.0217 - ETA: 39s - loss: 4.7407 - acc: 0.0218 - ETA: 38s - loss: 4.7413 - acc: 0.0217 - ETA: 37s - loss: 4.7415 - acc: 0.0220 - ETA: 36s - loss: 4.7418 - acc: 0.0219 - ETA: 35s - loss: 4.7416 - acc: 0.0222 - ETA: 34s - loss: 4.7420 - acc: 0.0221 - ETA: 33s - loss: 4.7421 - acc: 0.0222 - ETA: 33s - loss: 4.7426 - acc: 0.0221 - ETA: 32s - loss: 4.7427 - acc: 0.0222 - ETA: 31s - loss: 4.7431 - acc: 0.0223 - ETA: 30s - loss: 4.7432 - acc: 0.0224 - ETA: 29s - loss: 4.7430 - acc: 0.0224 - ETA: 28s - loss: 4.7430 - acc: 0.0223 - ETA: 27s - loss: 4.7433 - acc: 0.0222 - ETA: 26s - loss: 4.7430 - acc: 0.0221 - ETA: 25s - loss: 4.7429 - acc: 0.0221 - ETA: 24s - loss: 4.7429 - acc: 0.0220 - ETA: 23s - loss: 4.7439 - acc: 0.0221 - ETA: 22s - loss: 4.7443 - acc: 0.0220 - ETA: 22s - loss: 4.7446 - acc: 0.0219 - ETA: 21s - loss: 4.7447 - acc: 0.0219 - ETA: 20s - loss: 4.7448 - acc: 0.0218 - ETA: 19s - loss: 4.7446 - acc: 0.0217 - ETA: 18s - loss: 4.7452 - acc: 0.0217 - ETA: 17s - loss: 4.7448 - acc: 0.0216 - ETA: 16s - loss: 4.7450 - acc: 0.0217 - ETA: 15s - loss: 4.7444 - acc: 0.0219 - ETA: 14s - loss: 4.7447 - acc: 0.0220 - ETA: 13s - loss: 4.7447 - acc: 0.0221 - ETA: 12s - loss: 4.7450 - acc: 0.0220 - ETA: 11s - loss: 4.7448 - acc: 0.0220 - ETA: 11s - loss: 4.7443 - acc: 0.0220 - ETA: 10s - loss: 4.7444 - acc: 0.0223 - ETA: 9s - loss: 4.7439 - acc: 0.0225  - ETA: 8s - loss: 4.7439 - acc: 0.0225 - ETA: 7s - loss: 4.7437 - acc: 0.0224 - ETA: 6s - loss: 4.7435 - acc: 0.0223 - ETA: 5s - loss: 4.7437 - acc: 0.0223 - ETA: 4s - loss: 4.7448 - acc: 0.0223 - ETA: 3s - loss: 4.7449 - acc: 0.0223 - ETA: 2s - loss: 4.7447 - acc: 0.0224 - ETA: 1s - loss: 4.7450 - acc: 0.0223 - ETA: 0s - loss: 4.7450 - acc: 0.0222Epoch 00005: val_loss improved from 4.79618 to 4.76214, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 321s - loss: 4.7444 - acc: 0.0225 - val_loss: 4.7621 - val_acc: 0.0216
    Epoch 7/10
    6660/6680 [============================>.] - ETA: 303s - loss: 4.7271 - acc: 0.0000e+00 - ETA: 296s - loss: 4.7373 - acc: 0.0000e+00 - ETA: 295s - loss: 4.7169 - acc: 0.0000e+00 - ETA: 295s - loss: 4.7495 - acc: 0.0000e+00 - ETA: 295s - loss: 4.6926 - acc: 0.0100     - ETA: 293s - loss: 4.6810 - acc: 0.0167 - ETA: 292s - loss: 4.7219 - acc: 0.0214 - ETA: 291s - loss: 4.7172 - acc: 0.0313 - ETA: 289s - loss: 4.7293 - acc: 0.0278 - ETA: 288s - loss: 4.7239 - acc: 0.0250 - ETA: 288s - loss: 4.7109 - acc: 0.0227 - ETA: 287s - loss: 4.7035 - acc: 0.0208 - ETA: 286s - loss: 4.7109 - acc: 0.0231 - ETA: 286s - loss: 4.7202 - acc: 0.0214 - ETA: 285s - loss: 4.7224 - acc: 0.0233 - ETA: 285s - loss: 4.7086 - acc: 0.0250 - ETA: 283s - loss: 4.7151 - acc: 0.0235 - ETA: 283s - loss: 4.7086 - acc: 0.0250 - ETA: 282s - loss: 4.7157 - acc: 0.0237 - ETA: 281s - loss: 4.7197 - acc: 0.0250 - ETA: 281s - loss: 4.7250 - acc: 0.0238 - ETA: 279s - loss: 4.7243 - acc: 0.0273 - ETA: 278s - loss: 4.7256 - acc: 0.0326 - ETA: 277s - loss: 4.7263 - acc: 0.0313 - ETA: 277s - loss: 4.7305 - acc: 0.0320 - ETA: 275s - loss: 4.7270 - acc: 0.0327 - ETA: 275s - loss: 4.7268 - acc: 0.0352 - ETA: 274s - loss: 4.7254 - acc: 0.0339 - ETA: 273s - loss: 4.7265 - acc: 0.0328 - ETA: 272s - loss: 4.7301 - acc: 0.0333 - ETA: 271s - loss: 4.7343 - acc: 0.0323 - ETA: 270s - loss: 4.7287 - acc: 0.0328 - ETA: 269s - loss: 4.7318 - acc: 0.0333 - ETA: 268s - loss: 4.7308 - acc: 0.0338 - ETA: 267s - loss: 4.7311 - acc: 0.0357 - ETA: 266s - loss: 4.7308 - acc: 0.0347 - ETA: 265s - loss: 4.7257 - acc: 0.0351 - ETA: 264s - loss: 4.7308 - acc: 0.0342 - ETA: 264s - loss: 4.7338 - acc: 0.0346 - ETA: 263s - loss: 4.7401 - acc: 0.0338 - ETA: 262s - loss: 4.7405 - acc: 0.0341 - ETA: 261s - loss: 4.7395 - acc: 0.0333 - ETA: 260s - loss: 4.7412 - acc: 0.0349 - ETA: 259s - loss: 4.7399 - acc: 0.0375 - ETA: 258s - loss: 4.7415 - acc: 0.0367 - ETA: 257s - loss: 4.7402 - acc: 0.0370 - ETA: 256s - loss: 4.7372 - acc: 0.0362 - ETA: 256s - loss: 4.7338 - acc: 0.0354 - ETA: 255s - loss: 4.7327 - acc: 0.0357 - ETA: 254s - loss: 4.7339 - acc: 0.0350 - ETA: 253s - loss: 4.7304 - acc: 0.0373 - ETA: 252s - loss: 4.7346 - acc: 0.0365 - ETA: 251s - loss: 4.7402 - acc: 0.0358 - ETA: 250s - loss: 4.7382 - acc: 0.0361 - ETA: 249s - loss: 4.7394 - acc: 0.0355 - ETA: 248s - loss: 4.7399 - acc: 0.0348 - ETA: 247s - loss: 4.7442 - acc: 0.0342 - ETA: 247s - loss: 4.7465 - acc: 0.0336 - ETA: 246s - loss: 4.7498 - acc: 0.0331 - ETA: 245s - loss: 4.7488 - acc: 0.0333 - ETA: 244s - loss: 4.7503 - acc: 0.0328 - ETA: 243s - loss: 4.7475 - acc: 0.0323 - ETA: 242s - loss: 4.7477 - acc: 0.0317 - ETA: 241s - loss: 4.7444 - acc: 0.0313 - ETA: 240s - loss: 4.7434 - acc: 0.0315 - ETA: 239s - loss: 4.7431 - acc: 0.0311 - ETA: 238s - loss: 4.7413 - acc: 0.0306 - ETA: 238s - loss: 4.7405 - acc: 0.0301 - ETA: 237s - loss: 4.7398 - acc: 0.0297 - ETA: 236s - loss: 4.7405 - acc: 0.0300 - ETA: 235s - loss: 4.7395 - acc: 0.0303 - ETA: 234s - loss: 4.7372 - acc: 0.0299 - ETA: 233s - loss: 4.7385 - acc: 0.0295 - ETA: 232s - loss: 4.7418 - acc: 0.0291 - ETA: 231s - loss: 4.7405 - acc: 0.0300 - ETA: 231s - loss: 4.7393 - acc: 0.0303 - ETA: 230s - loss: 4.7413 - acc: 0.0299 - ETA: 229s - loss: 4.7414 - acc: 0.0295 - ETA: 228s - loss: 4.7434 - acc: 0.0291 - ETA: 227s - loss: 4.7400 - acc: 0.0288 - ETA: 226s - loss: 4.7420 - acc: 0.0284 - ETA: 225s - loss: 4.7425 - acc: 0.0280 - ETA: 224s - loss: 4.7397 - acc: 0.0283 - ETA: 223s - loss: 4.7398 - acc: 0.0298 - ETA: 222s - loss: 4.7379 - acc: 0.0300 - ETA: 221s - loss: 4.7357 - acc: 0.0308 - ETA: 221s - loss: 4.7340 - acc: 0.0305 - ETA: 220s - loss: 4.7337 - acc: 0.0301 - ETA: 219s - loss: 4.7327 - acc: 0.0298 - ETA: 218s - loss: 4.7312 - acc: 0.0300 - ETA: 217s - loss: 4.7303 - acc: 0.0297 - ETA: 216s - loss: 4.7316 - acc: 0.0293 - ETA: 215s - loss: 4.7331 - acc: 0.0290 - ETA: 215s - loss: 4.7316 - acc: 0.0293 - ETA: 214s - loss: 4.7312 - acc: 0.0295 - ETA: 213s - loss: 4.7320 - acc: 0.0292 - ETA: 212s - loss: 4.7329 - acc: 0.0289 - ETA: 211s - loss: 4.7340 - acc: 0.0286 - ETA: 210s - loss: 4.7324 - acc: 0.0288 - ETA: 209s - loss: 4.7333 - acc: 0.0290 - ETA: 208s - loss: 4.7317 - acc: 0.0287 - ETA: 207s - loss: 4.7330 - acc: 0.0284 - ETA: 206s - loss: 4.7346 - acc: 0.0282 - ETA: 206s - loss: 4.7325 - acc: 0.0284 - ETA: 205s - loss: 4.7333 - acc: 0.0281 - ETA: 204s - loss: 4.7315 - acc: 0.0278 - ETA: 203s - loss: 4.7303 - acc: 0.0280 - ETA: 202s - loss: 4.7297 - acc: 0.0287 - ETA: 201s - loss: 4.7284 - acc: 0.0284 - ETA: 200s - loss: 4.7275 - acc: 0.0282 - ETA: 199s - loss: 4.7272 - acc: 0.0288 - ETA: 198s - loss: 4.7279 - acc: 0.0290 - ETA: 197s - loss: 4.7276 - acc: 0.0288 - ETA: 196s - loss: 4.7270 - acc: 0.0285 - ETA: 196s - loss: 4.7274 - acc: 0.0287 - ETA: 195s - loss: 4.7275 - acc: 0.0284 - ETA: 194s - loss: 4.7262 - acc: 0.0282 - ETA: 193s - loss: 4.7259 - acc: 0.0288 - ETA: 192s - loss: 4.7252 - acc: 0.0286 - ETA: 191s - loss: 4.7230 - acc: 0.0288 - ETA: 190s - loss: 4.7222 - acc: 0.0285 - ETA: 189s - loss: 4.7225 - acc: 0.0283 - ETA: 188s - loss: 4.7229 - acc: 0.0280 - ETA: 187s - loss: 4.7243 - acc: 0.0278 - ETA: 186s - loss: 4.7240 - acc: 0.0276 - ETA: 186s - loss: 4.7226 - acc: 0.0278 - ETA: 185s - loss: 4.7209 - acc: 0.0280 - ETA: 184s - loss: 4.7202 - acc: 0.0281 - ETA: 183s - loss: 4.7183 - acc: 0.0283 - ETA: 182s - loss: 4.7176 - acc: 0.0281 - ETA: 181s - loss: 4.7155 - acc: 0.0282 - ETA: 180s - loss: 4.7145 - acc: 0.0280 - ETA: 179s - loss: 4.7159 - acc: 0.0278 - ETA: 178s - loss: 4.7173 - acc: 0.0276 - ETA: 177s - loss: 4.7174 - acc: 0.0274 - ETA: 176s - loss: 4.7169 - acc: 0.0272 - ETA: 176s - loss: 4.7168 - acc: 0.0274 - ETA: 175s - loss: 4.7171 - acc: 0.0272 - ETA: 174s - loss: 4.7174 - acc: 0.0270 - ETA: 173s - loss: 4.7178 - acc: 0.0279 - ETA: 172s - loss: 4.7181 - acc: 0.0280 - ETA: 171s - loss: 4.7179 - acc: 0.0278 - ETA: 170s - loss: 4.7180 - acc: 0.0280 - ETA: 169s - loss: 4.7188 - acc: 0.0278 - ETA: 168s - loss: 4.7192 - acc: 0.0283 - ETA: 167s - loss: 4.7197 - acc: 0.0284 - ETA: 167s - loss: 4.7192 - acc: 0.0282 - ETA: 166s - loss: 4.7183 - acc: 0.0284 - ETA: 165s - loss: 4.7173 - acc: 0.0285 - ETA: 164s - loss: 4.7178 - acc: 0.0283 - ETA: 163s - loss: 4.7180 - acc: 0.0281 - ETA: 162s - loss: 4.7198 - acc: 0.0280 - ETA: 161s - loss: 4.7183 - acc: 0.0281 - ETA: 160s - loss: 4.7189 - acc: 0.0282 - ETA: 159s - loss: 4.7188 - acc: 0.0284 - ETA: 159s - loss: 4.7195 - acc: 0.0282 - ETA: 158s - loss: 4.7194 - acc: 0.0280 - ETA: 157s - loss: 4.7196 - acc: 0.0278 - ETA: 156s - loss: 4.7192 - acc: 0.0280 - ETA: 155s - loss: 4.7189 - acc: 0.0278 - ETA: 154s - loss: 4.7186 - acc: 0.0283 - ETA: 153s - loss: 4.7178 - acc: 0.0281 - ETA: 152s - loss: 4.7173 - acc: 0.0282 - ETA: 151s - loss: 4.7166 - acc: 0.0287 - ETA: 151s - loss: 4.7173 - acc: 0.0288 - ETA: 150s - loss: 4.7169 - acc: 0.0289 - ETA: 149s - loss: 4.7162 - acc: 0.0287 - ETA: 148s - loss: 4.7160 - acc: 0.0286 - ETA: 147s - loss: 4.7150 - acc: 0.0287 - ETA: 146s - loss: 4.7165 - acc: 0.0285 - ETA: 145s - loss: 4.7187 - acc: 0.0284 - ETA: 144s - loss: 4.7180 - acc: 0.0288 - ETA: 143s - loss: 4.7177 - acc: 0.0286 - ETA: 142s - loss: 4.7186 - acc: 0.0287 - ETA: 142s - loss: 4.7183 - acc: 0.0289 - ETA: 141s - loss: 4.7186 - acc: 0.0287 - ETA: 140s - loss: 4.7179 - acc: 0.0285 - ETA: 139s - loss: 4.7172 - acc: 0.0287 - ETA: 138s - loss: 4.7170 - acc: 0.0285 - ETA: 137s - loss: 4.7183 - acc: 0.0283 - ETA: 136s - loss: 4.7187 - acc: 0.0282 - ETA: 135s - loss: 4.7180 - acc: 0.0280 - ETA: 134s - loss: 4.7193 - acc: 0.0279 - ETA: 134s - loss: 4.7194 - acc: 0.0280 - ETA: 133s - loss: 4.7190 - acc: 0.0278 - ETA: 132s - loss: 4.7190 - acc: 0.0277 - ETA: 131s - loss: 4.7178 - acc: 0.0278 - ETA: 130s - loss: 4.7179 - acc: 0.0282 - ETA: 129s - loss: 4.7179 - acc: 0.0280 - ETA: 128s - loss: 4.7169 - acc: 0.0282 - ETA: 127s - loss: 4.7160 - acc: 0.0280 - ETA: 126s - loss: 4.7158 - acc: 0.0284 - ETA: 126s - loss: 4.7171 - acc: 0.0282 - ETA: 125s - loss: 4.7170 - acc: 0.0281 - ETA: 124s - loss: 4.7175 - acc: 0.0279 - ETA: 123s - loss: 4.7169 - acc: 0.0278 - ETA: 122s - loss: 4.7164 - acc: 0.0277 - ETA: 121s - loss: 4.7182 - acc: 0.0278 - ETA: 120s - loss: 4.7176 - acc: 0.0276 - ETA: 119s - loss: 4.7170 - acc: 0.0275 - ETA: 118s - loss: 4.7169 - acc: 0.0279 - ETA: 117s - loss: 4.7162 - acc: 0.0277 - ETA: 117s - loss: 4.7153 - acc: 0.0278 - ETA: 116s - loss: 4.7161 - acc: 0.0277 - ETA: 115s - loss: 4.7171 - acc: 0.0276 - ETA: 114s - loss: 4.7175 - acc: 0.0274 - ETA: 113s - loss: 4.7177 - acc: 0.0273 - ETA: 112s - loss: 4.7173 - acc: 0.0272 - ETA: 111s - loss: 4.7180 - acc: 0.0273 - ETA: 110s - loss: 4.7172 - acc: 0.0274 - ETA: 109s - loss: 4.7180 - acc: 0.0277 - ETA: 108s - loss: 4.7173 - acc: 0.0278 - ETA: 108s - loss: 4.7176 - acc: 0.0277 - ETA: 107s - loss: 4.7179 - acc: 0.0280 - ETA: 106s - loss: 4.7171 - acc: 0.0279 - ETA: 105s - loss: 4.7168 - acc: 0.0278 - ETA: 104s - loss: 4.7170 - acc: 0.0279 - ETA: 103s - loss: 4.7169 - acc: 0.0278 - ETA: 102s - loss: 4.7170 - acc: 0.0276 - ETA: 101s - loss: 4.7170 - acc: 0.0275 - ETA: 100s - loss: 4.7166 - acc: 0.0278 - ETA: 100s - loss: 4.7171 - acc: 0.0277 - ETA: 99s - loss: 4.7175 - acc: 0.0276  - ETA: 98s - loss: 4.7168 - acc: 0.0275 - ETA: 97s - loss: 4.7158 - acc: 0.0276 - ETA: 96s - loss: 4.7162 - acc: 0.0274 - ETA: 95s - loss: 4.7167 - acc: 0.0273 - ETA: 94s - loss: 4.7167 - acc: 0.0274 - ETA: 93s - loss: 4.7164 - acc: 0.0275 - ETA: 92s - loss: 4.7169 - acc: 0.0274 - ETA: 92s - loss: 4.7170 - acc: 0.0275 - ETA: 91s - loss: 4.7168 - acc: 0.0274 - ETA: 90s - loss: 4.7165 - acc: 0.0277 - ETA: 89s - loss: 4.7166 - acc: 0.0276 - ETA: 88s - loss: 4.7164 - acc: 0.0277 - ETA: 87s - loss: 4.7161 - acc: 0.0275 - ETA: 86s - loss: 4.7162 - acc: 0.0276 - ETA: 85s - loss: 4.7167 - acc: 0.0275 - ETA: 84s - loss: 4.7180 - acc: 0.0274 - ETA: 84s - loss: 4.7178 - acc: 0.0275 - ETA: 83s - loss: 4.7179 - acc: 0.0274 - ETA: 82s - loss: 4.7182 - acc: 0.0275 - ETA: 81s - loss: 4.7185 - acc: 0.0274 - ETA: 80s - loss: 4.7179 - acc: 0.0275 - ETA: 79s - loss: 4.7177 - acc: 0.0273 - ETA: 78s - loss: 4.7180 - acc: 0.0272 - ETA: 77s - loss: 4.7173 - acc: 0.0271 - ETA: 76s - loss: 4.7174 - acc: 0.0272 - ETA: 75s - loss: 4.7168 - acc: 0.0273 - ETA: 75s - loss: 4.7162 - acc: 0.0272 - ETA: 74s - loss: 4.7165 - acc: 0.0273 - ETA: 73s - loss: 4.7158 - acc: 0.0274 - ETA: 72s - loss: 4.7157 - acc: 0.0275 - ETA: 71s - loss: 4.7157 - acc: 0.0276 - ETA: 70s - loss: 4.7160 - acc: 0.0280 - ETA: 69s - loss: 4.7160 - acc: 0.0279 - ETA: 68s - loss: 4.7163 - acc: 0.0280 - ETA: 67s - loss: 4.7156 - acc: 0.0283 - ETA: 67s - loss: 4.7151 - acc: 0.0282 - ETA: 66s - loss: 4.7155 - acc: 0.0281 - ETA: 65s - loss: 4.7158 - acc: 0.0284 - ETA: 64s - loss: 4.7159 - acc: 0.0282 - ETA: 63s - loss: 4.7160 - acc: 0.0281 - ETA: 62s - loss: 4.7164 - acc: 0.0280 - ETA: 61s - loss: 4.7172 - acc: 0.0279 - ETA: 60s - loss: 4.7175 - acc: 0.0280 - ETA: 59s - loss: 4.7174 - acc: 0.0281 - ETA: 58s - loss: 4.7165 - acc: 0.0285 - ETA: 58s - loss: 4.7169 - acc: 0.0286 - ETA: 57s - loss: 4.7168 - acc: 0.0289 - ETA: 56s - loss: 4.7165 - acc: 0.0288 - ETA: 55s - loss: 4.7157 - acc: 0.0287 - ETA: 54s - loss: 4.7152 - acc: 0.0286 - ETA: 53s - loss: 4.7149 - acc: 0.0286 - ETA: 52s - loss: 4.7154 - acc: 0.0285 - ETA: 51s - loss: 4.7145 - acc: 0.0286 - ETA: 50s - loss: 4.7142 - acc: 0.0285 - ETA: 50s - loss: 4.7136 - acc: 0.0284 - ETA: 49s - loss: 4.7131 - acc: 0.0283 - ETA: 48s - loss: 4.7138 - acc: 0.0282 - ETA: 47s - loss: 4.7141 - acc: 0.0283 - ETA: 46s - loss: 4.7144 - acc: 0.0284 - ETA: 45s - loss: 4.7142 - acc: 0.0284 - ETA: 44s - loss: 4.7144 - acc: 0.0287 - ETA: 43s - loss: 4.7137 - acc: 0.0286 - ETA: 42s - loss: 4.7137 - acc: 0.0288 - ETA: 42s - loss: 4.7138 - acc: 0.0287 - ETA: 41s - loss: 4.7133 - acc: 0.0288 - ETA: 40s - loss: 4.7138 - acc: 0.0287 - ETA: 39s - loss: 4.7138 - acc: 0.0286 - ETA: 38s - loss: 4.7137 - acc: 0.0287 - ETA: 37s - loss: 4.7125 - acc: 0.0289 - ETA: 36s - loss: 4.7126 - acc: 0.0290 - ETA: 35s - loss: 4.7130 - acc: 0.0289 - ETA: 34s - loss: 4.7129 - acc: 0.0290 - ETA: 33s - loss: 4.7126 - acc: 0.0289 - ETA: 33s - loss: 4.7121 - acc: 0.0288 - ETA: 32s - loss: 4.7122 - acc: 0.0289 - ETA: 31s - loss: 4.7115 - acc: 0.0289 - ETA: 30s - loss: 4.7132 - acc: 0.0288 - ETA: 29s - loss: 4.7134 - acc: 0.0287 - ETA: 28s - loss: 4.7134 - acc: 0.0286 - ETA: 27s - loss: 4.7135 - acc: 0.0285 - ETA: 26s - loss: 4.7136 - acc: 0.0285 - ETA: 25s - loss: 4.7142 - acc: 0.0284 - ETA: 25s - loss: 4.7148 - acc: 0.0284 - ETA: 24s - loss: 4.7156 - acc: 0.0283 - ETA: 23s - loss: 4.7149 - acc: 0.0282 - ETA: 22s - loss: 4.7150 - acc: 0.0282 - ETA: 21s - loss: 4.7150 - acc: 0.0281 - ETA: 20s - loss: 4.7146 - acc: 0.0280 - ETA: 19s - loss: 4.7146 - acc: 0.0279 - ETA: 18s - loss: 4.7142 - acc: 0.0278 - ETA: 17s - loss: 4.7135 - acc: 0.0282 - ETA: 16s - loss: 4.7130 - acc: 0.0281 - ETA: 16s - loss: 4.7127 - acc: 0.0282 - ETA: 15s - loss: 4.7125 - acc: 0.0282 - ETA: 14s - loss: 4.7131 - acc: 0.0281 - ETA: 13s - loss: 4.7129 - acc: 0.0281 - ETA: 12s - loss: 4.7128 - acc: 0.0280 - ETA: 11s - loss: 4.7127 - acc: 0.0283 - ETA: 10s - loss: 4.7130 - acc: 0.0283 - ETA: 9s - loss: 4.7132 - acc: 0.0282  - ETA: 8s - loss: 4.7128 - acc: 0.0281 - ETA: 8s - loss: 4.7134 - acc: 0.0280 - ETA: 7s - loss: 4.7138 - acc: 0.0279 - ETA: 6s - loss: 4.7134 - acc: 0.0278 - ETA: 5s - loss: 4.7124 - acc: 0.0279 - ETA: 4s - loss: 4.7126 - acc: 0.0278 - ETA: 3s - loss: 4.7132 - acc: 0.0277 - ETA: 2s - loss: 4.7126 - acc: 0.0276 - ETA: 1s - loss: 4.7129 - acc: 0.0276 - ETA: 0s - loss: 4.7126 - acc: 0.0275Epoch 00006: val_loss improved from 4.76214 to 4.74435, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 314s - loss: 4.7124 - acc: 0.0275 - val_loss: 4.7443 - val_acc: 0.0240
    Epoch 8/10
    6660/6680 [============================>.] - ETA: 295s - loss: 4.9397 - acc: 0.0000e+00 - ETA: 297s - loss: 4.8384 - acc: 0.0000e+00 - ETA: 296s - loss: 4.6922 - acc: 0.0000e+00 - ETA: 296s - loss: 4.7290 - acc: 0.0000e+00 - ETA: 294s - loss: 4.7374 - acc: 0.0000e+00 - ETA: 293s - loss: 4.7341 - acc: 0.0083     - ETA: 292s - loss: 4.7205 - acc: 0.0071 - ETA: 291s - loss: 4.7221 - acc: 0.0125 - ETA: 291s - loss: 4.7251 - acc: 0.0111 - ETA: 290s - loss: 4.7261 - acc: 0.0150 - ETA: 290s - loss: 4.7004 - acc: 0.0182 - ETA: 288s - loss: 4.6941 - acc: 0.0208 - ETA: 287s - loss: 4.7065 - acc: 0.0192 - ETA: 286s - loss: 4.6943 - acc: 0.0250 - ETA: 285s - loss: 4.6845 - acc: 0.0267 - ETA: 284s - loss: 4.6845 - acc: 0.0281 - ETA: 283s - loss: 4.6839 - acc: 0.0324 - ETA: 282s - loss: 4.6941 - acc: 0.0306 - ETA: 282s - loss: 4.7056 - acc: 0.0289 - ETA: 281s - loss: 4.7004 - acc: 0.0300 - ETA: 280s - loss: 4.6856 - acc: 0.0357 - ETA: 280s - loss: 4.6816 - acc: 0.0341 - ETA: 279s - loss: 4.6788 - acc: 0.0326 - ETA: 278s - loss: 4.6810 - acc: 0.0313 - ETA: 277s - loss: 4.6850 - acc: 0.0300 - ETA: 276s - loss: 4.6774 - acc: 0.0308 - ETA: 275s - loss: 4.6734 - acc: 0.0315 - ETA: 274s - loss: 4.6715 - acc: 0.0321 - ETA: 273s - loss: 4.6697 - acc: 0.0310 - ETA: 272s - loss: 4.6659 - acc: 0.0300 - ETA: 271s - loss: 4.6632 - acc: 0.0306 - ETA: 270s - loss: 4.6628 - acc: 0.0313 - ETA: 270s - loss: 4.6634 - acc: 0.0303 - ETA: 269s - loss: 4.6593 - acc: 0.0309 - ETA: 268s - loss: 4.6614 - acc: 0.0300 - ETA: 267s - loss: 4.6600 - acc: 0.0306 - ETA: 266s - loss: 4.6617 - acc: 0.0311 - ETA: 265s - loss: 4.6594 - acc: 0.0329 - ETA: 264s - loss: 4.6622 - acc: 0.0333 - ETA: 263s - loss: 4.6635 - acc: 0.0350 - ETA: 262s - loss: 4.6683 - acc: 0.0354 - ETA: 261s - loss: 4.6674 - acc: 0.0369 - ETA: 261s - loss: 4.6704 - acc: 0.0360 - ETA: 260s - loss: 4.6677 - acc: 0.0375 - ETA: 259s - loss: 4.6659 - acc: 0.0378 - ETA: 258s - loss: 4.6673 - acc: 0.0380 - ETA: 257s - loss: 4.6664 - acc: 0.0383 - ETA: 256s - loss: 4.6747 - acc: 0.0375 - ETA: 256s - loss: 4.6733 - acc: 0.0367 - ETA: 255s - loss: 4.6723 - acc: 0.0380 - ETA: 254s - loss: 4.6740 - acc: 0.0373 - ETA: 253s - loss: 4.6771 - acc: 0.0365 - ETA: 252s - loss: 4.6798 - acc: 0.0358 - ETA: 251s - loss: 4.6825 - acc: 0.0352 - ETA: 250s - loss: 4.6850 - acc: 0.0355 - ETA: 249s - loss: 4.6837 - acc: 0.0348 - ETA: 248s - loss: 4.6861 - acc: 0.0342 - ETA: 247s - loss: 4.6874 - acc: 0.0336 - ETA: 246s - loss: 4.6873 - acc: 0.0331 - ETA: 245s - loss: 4.6876 - acc: 0.0333 - ETA: 245s - loss: 4.6881 - acc: 0.0336 - ETA: 244s - loss: 4.6876 - acc: 0.0347 - ETA: 243s - loss: 4.6910 - acc: 0.0341 - ETA: 242s - loss: 4.6892 - acc: 0.0336 - ETA: 241s - loss: 4.6918 - acc: 0.0331 - ETA: 240s - loss: 4.6897 - acc: 0.0326 - ETA: 239s - loss: 4.6889 - acc: 0.0321 - ETA: 238s - loss: 4.6907 - acc: 0.0316 - ETA: 237s - loss: 4.6922 - acc: 0.0319 - ETA: 237s - loss: 4.6948 - acc: 0.0314 - ETA: 236s - loss: 4.6949 - acc: 0.0310 - ETA: 235s - loss: 4.6944 - acc: 0.0306 - ETA: 234s - loss: 4.6916 - acc: 0.0308 - ETA: 233s - loss: 4.6879 - acc: 0.0311 - ETA: 232s - loss: 4.6855 - acc: 0.0320 - ETA: 231s - loss: 4.6815 - acc: 0.0329 - ETA: 230s - loss: 4.6818 - acc: 0.0338 - ETA: 230s - loss: 4.6832 - acc: 0.0333 - ETA: 229s - loss: 4.6818 - acc: 0.0348 - ETA: 228s - loss: 4.6816 - acc: 0.0356 - ETA: 227s - loss: 4.6826 - acc: 0.0352 - ETA: 226s - loss: 4.6851 - acc: 0.0348 - ETA: 225s - loss: 4.6844 - acc: 0.0343 - ETA: 224s - loss: 4.6836 - acc: 0.0339 - ETA: 223s - loss: 4.6851 - acc: 0.0341 - ETA: 222s - loss: 4.6836 - acc: 0.0343 - ETA: 222s - loss: 4.6850 - acc: 0.0345 - ETA: 221s - loss: 4.6857 - acc: 0.0341 - ETA: 220s - loss: 4.6857 - acc: 0.0337 - ETA: 219s - loss: 4.6851 - acc: 0.0339 - ETA: 218s - loss: 4.6842 - acc: 0.0335 - ETA: 217s - loss: 4.6826 - acc: 0.0348 - ETA: 216s - loss: 4.6822 - acc: 0.0344 - ETA: 215s - loss: 4.6830 - acc: 0.0340 - ETA: 215s - loss: 4.6819 - acc: 0.0337 - ETA: 214s - loss: 4.6807 - acc: 0.0333 - ETA: 213s - loss: 4.6777 - acc: 0.0330 - ETA: 212s - loss: 4.6828 - acc: 0.0332 - ETA: 211s - loss: 4.6820 - acc: 0.0328 - ETA: 210s - loss: 4.6812 - acc: 0.0325 - ETA: 209s - loss: 4.6817 - acc: 0.0327 - ETA: 208s - loss: 4.6797 - acc: 0.0328 - ETA: 207s - loss: 4.6789 - acc: 0.0325 - ETA: 206s - loss: 4.6808 - acc: 0.0322 - ETA: 205s - loss: 4.6834 - acc: 0.0319 - ETA: 205s - loss: 4.6839 - acc: 0.0321 - ETA: 204s - loss: 4.6845 - acc: 0.0322 - ETA: 203s - loss: 4.6847 - acc: 0.0319 - ETA: 202s - loss: 4.6852 - acc: 0.0317 - ETA: 201s - loss: 4.6853 - acc: 0.0323 - ETA: 200s - loss: 4.6855 - acc: 0.0320 - ETA: 199s - loss: 4.6858 - acc: 0.0317 - ETA: 198s - loss: 4.6866 - acc: 0.0314 - ETA: 197s - loss: 4.6861 - acc: 0.0316 - ETA: 197s - loss: 4.6870 - acc: 0.0313 - ETA: 196s - loss: 4.6880 - acc: 0.0310 - ETA: 195s - loss: 4.6875 - acc: 0.0312 - ETA: 194s - loss: 4.6876 - acc: 0.0318 - ETA: 193s - loss: 4.6876 - acc: 0.0315 - ETA: 192s - loss: 4.6883 - acc: 0.0313 - ETA: 191s - loss: 4.6879 - acc: 0.0310 - ETA: 190s - loss: 4.6885 - acc: 0.0311 - ETA: 189s - loss: 4.6887 - acc: 0.0313 - ETA: 188s - loss: 4.6888 - acc: 0.0315 - ETA: 188s - loss: 4.6918 - acc: 0.0312 - ETA: 187s - loss: 4.6922 - acc: 0.0310 - ETA: 186s - loss: 4.6932 - acc: 0.0307 - ETA: 185s - loss: 4.6940 - acc: 0.0305 - ETA: 184s - loss: 4.6939 - acc: 0.0306 - ETA: 183s - loss: 4.6931 - acc: 0.0308 - ETA: 182s - loss: 4.6932 - acc: 0.0305 - ETA: 181s - loss: 4.6924 - acc: 0.0303 - ETA: 180s - loss: 4.6920 - acc: 0.0308 - ETA: 180s - loss: 4.6904 - acc: 0.0310 - ETA: 179s - loss: 4.6919 - acc: 0.0311 - ETA: 178s - loss: 4.6915 - acc: 0.0309 - ETA: 177s - loss: 4.6892 - acc: 0.0314 - ETA: 176s - loss: 4.6881 - acc: 0.0315 - ETA: 175s - loss: 4.6893 - acc: 0.0313 - ETA: 174s - loss: 4.6890 - acc: 0.0314 - ETA: 173s - loss: 4.6897 - acc: 0.0319 - ETA: 172s - loss: 4.6882 - acc: 0.0317 - ETA: 171s - loss: 4.6884 - acc: 0.0315 - ETA: 171s - loss: 4.6889 - acc: 0.0313 - ETA: 170s - loss: 4.6905 - acc: 0.0310 - ETA: 169s - loss: 4.6888 - acc: 0.0312 - ETA: 168s - loss: 4.6882 - acc: 0.0313 - ETA: 167s - loss: 4.6886 - acc: 0.0314 - ETA: 166s - loss: 4.6902 - acc: 0.0312 - ETA: 165s - loss: 4.6898 - acc: 0.0310 - ETA: 164s - loss: 4.6890 - acc: 0.0308 - ETA: 163s - loss: 4.6900 - acc: 0.0309 - ETA: 162s - loss: 4.6904 - acc: 0.0307 - ETA: 162s - loss: 4.6894 - acc: 0.0308 - ETA: 161s - loss: 4.6891 - acc: 0.0306 - ETA: 160s - loss: 4.6884 - acc: 0.0304 - ETA: 159s - loss: 4.6883 - acc: 0.0309 - ETA: 158s - loss: 4.6873 - acc: 0.0310 - ETA: 157s - loss: 4.6860 - acc: 0.0308 - ETA: 156s - loss: 4.6837 - acc: 0.0316 - ETA: 155s - loss: 4.6830 - acc: 0.0320 - ETA: 154s - loss: 4.6835 - acc: 0.0318 - ETA: 153s - loss: 4.6839 - acc: 0.0316 - ETA: 153s - loss: 4.6832 - acc: 0.0317 - ETA: 152s - loss: 4.6824 - acc: 0.0315 - ETA: 151s - loss: 4.6824 - acc: 0.0319 - ETA: 150s - loss: 4.6827 - acc: 0.0320 - ETA: 149s - loss: 4.6813 - acc: 0.0324 - ETA: 148s - loss: 4.6811 - acc: 0.0325 - ETA: 147s - loss: 4.6801 - acc: 0.0326 - ETA: 146s - loss: 4.6812 - acc: 0.0325 - ETA: 145s - loss: 4.6830 - acc: 0.0323 - ETA: 144s - loss: 4.6834 - acc: 0.0321 - ETA: 143s - loss: 4.6830 - acc: 0.0322 - ETA: 143s - loss: 4.6836 - acc: 0.0323 - ETA: 142s - loss: 4.6826 - acc: 0.0321 - ETA: 141s - loss: 4.6829 - acc: 0.0319 - ETA: 140s - loss: 4.6829 - acc: 0.0320 - ETA: 139s - loss: 4.6825 - acc: 0.0318 - ETA: 138s - loss: 4.6829 - acc: 0.0319 - ETA: 137s - loss: 4.6844 - acc: 0.0323 - ETA: 136s - loss: 4.6839 - acc: 0.0324 - ETA: 135s - loss: 4.6830 - acc: 0.0325 - ETA: 134s - loss: 4.6822 - acc: 0.0326 - ETA: 134s - loss: 4.6822 - acc: 0.0327 - ETA: 133s - loss: 4.6814 - acc: 0.0325 - ETA: 132s - loss: 4.6825 - acc: 0.0326 - ETA: 131s - loss: 4.6820 - acc: 0.0327 - ETA: 130s - loss: 4.6818 - acc: 0.0325 - ETA: 129s - loss: 4.6818 - acc: 0.0326 - ETA: 128s - loss: 4.6809 - acc: 0.0325 - ETA: 127s - loss: 4.6806 - acc: 0.0323 - ETA: 126s - loss: 4.6811 - acc: 0.0321 - ETA: 125s - loss: 4.6804 - acc: 0.0320 - ETA: 125s - loss: 4.6805 - acc: 0.0321 - ETA: 124s - loss: 4.6794 - acc: 0.0319 - ETA: 123s - loss: 4.6786 - acc: 0.0317 - ETA: 122s - loss: 4.6791 - acc: 0.0316 - ETA: 121s - loss: 4.6781 - acc: 0.0319 - ETA: 120s - loss: 4.6777 - acc: 0.0318 - ETA: 119s - loss: 4.6785 - acc: 0.0316 - ETA: 118s - loss: 4.6781 - acc: 0.0319 - ETA: 117s - loss: 4.6778 - acc: 0.0318 - ETA: 116s - loss: 4.6800 - acc: 0.0316 - ETA: 116s - loss: 4.6800 - acc: 0.0315 - ETA: 115s - loss: 4.6802 - acc: 0.0313 - ETA: 114s - loss: 4.6803 - acc: 0.0312 - ETA: 113s - loss: 4.6802 - acc: 0.0310 - ETA: 112s - loss: 4.6811 - acc: 0.0311 - ETA: 111s - loss: 4.6810 - acc: 0.0310 - ETA: 110s - loss: 4.6803 - acc: 0.0308 - ETA: 109s - loss: 4.6795 - acc: 0.0309 - ETA: 108s - loss: 4.6808 - acc: 0.0308 - ETA: 107s - loss: 4.6802 - acc: 0.0306 - ETA: 107s - loss: 4.6803 - acc: 0.0305 - ETA: 106s - loss: 4.6808 - acc: 0.0303 - ETA: 105s - loss: 4.6822 - acc: 0.0302 - ETA: 104s - loss: 4.6827 - acc: 0.0300 - ETA: 103s - loss: 4.6830 - acc: 0.0299 - ETA: 102s - loss: 4.6827 - acc: 0.0300 - ETA: 101s - loss: 4.6815 - acc: 0.0303 - ETA: 100s - loss: 4.6802 - acc: 0.0302 - ETA: 99s - loss: 4.6802 - acc: 0.0303  - ETA: 99s - loss: 4.6799 - acc: 0.0304 - ETA: 98s - loss: 4.6797 - acc: 0.0302 - ETA: 97s - loss: 4.6807 - acc: 0.0301 - ETA: 96s - loss: 4.6808 - acc: 0.0302 - ETA: 95s - loss: 4.6805 - acc: 0.0300 - ETA: 94s - loss: 4.6810 - acc: 0.0299 - ETA: 93s - loss: 4.6799 - acc: 0.0300 - ETA: 92s - loss: 4.6811 - acc: 0.0299 - ETA: 91s - loss: 4.6818 - acc: 0.0297 - ETA: 90s - loss: 4.6820 - acc: 0.0296 - ETA: 90s - loss: 4.6828 - acc: 0.0299 - ETA: 89s - loss: 4.6834 - acc: 0.0298 - ETA: 88s - loss: 4.6831 - acc: 0.0297 - ETA: 87s - loss: 4.6833 - acc: 0.0297 - ETA: 86s - loss: 4.6829 - acc: 0.0300 - ETA: 85s - loss: 4.6822 - acc: 0.0301 - ETA: 84s - loss: 4.6818 - acc: 0.0300 - ETA: 83s - loss: 4.6829 - acc: 0.0299 - ETA: 82s - loss: 4.6832 - acc: 0.0298 - ETA: 82s - loss: 4.6839 - acc: 0.0296 - ETA: 81s - loss: 4.6841 - acc: 0.0295 - ETA: 80s - loss: 4.6835 - acc: 0.0296 - ETA: 79s - loss: 4.6847 - acc: 0.0295 - ETA: 78s - loss: 4.6854 - acc: 0.0294 - ETA: 77s - loss: 4.6858 - acc: 0.0292 - ETA: 76s - loss: 4.6859 - acc: 0.0291 - ETA: 75s - loss: 4.6856 - acc: 0.0290 - ETA: 74s - loss: 4.6854 - acc: 0.0289 - ETA: 73s - loss: 4.6856 - acc: 0.0288 - ETA: 73s - loss: 4.6862 - acc: 0.0287 - ETA: 72s - loss: 4.6865 - acc: 0.0289 - ETA: 71s - loss: 4.6862 - acc: 0.0288 - ETA: 70s - loss: 4.6860 - acc: 0.0287 - ETA: 69s - loss: 4.6858 - acc: 0.0286 - ETA: 68s - loss: 4.6858 - acc: 0.0285 - ETA: 67s - loss: 4.6858 - acc: 0.0284 - ETA: 66s - loss: 4.6862 - acc: 0.0285 - ETA: 65s - loss: 4.6861 - acc: 0.0285 - ETA: 64s - loss: 4.6857 - acc: 0.0286 - ETA: 64s - loss: 4.6863 - acc: 0.0285 - ETA: 63s - loss: 4.6858 - acc: 0.0288 - ETA: 62s - loss: 4.6853 - acc: 0.0289 - ETA: 61s - loss: 4.6852 - acc: 0.0289 - ETA: 60s - loss: 4.6850 - acc: 0.0288 - ETA: 59s - loss: 4.6852 - acc: 0.0289 - ETA: 58s - loss: 4.6857 - acc: 0.0288 - ETA: 57s - loss: 4.6851 - acc: 0.0291 - ETA: 56s - loss: 4.6851 - acc: 0.0290 - ETA: 55s - loss: 4.6850 - acc: 0.0289 - ETA: 55s - loss: 4.6853 - acc: 0.0289 - ETA: 54s - loss: 4.6851 - acc: 0.0290 - ETA: 53s - loss: 4.6849 - acc: 0.0293 - ETA: 52s - loss: 4.6843 - acc: 0.0292 - ETA: 51s - loss: 4.6845 - acc: 0.0291 - ETA: 50s - loss: 4.6836 - acc: 0.0293 - ETA: 49s - loss: 4.6835 - acc: 0.0294 - ETA: 48s - loss: 4.6836 - acc: 0.0295 - ETA: 47s - loss: 4.6833 - acc: 0.0294 - ETA: 46s - loss: 4.6842 - acc: 0.0293 - ETA: 46s - loss: 4.6845 - acc: 0.0293 - ETA: 45s - loss: 4.6842 - acc: 0.0292 - ETA: 44s - loss: 4.6841 - acc: 0.0291 - ETA: 43s - loss: 4.6843 - acc: 0.0290 - ETA: 42s - loss: 4.6847 - acc: 0.0291 - ETA: 41s - loss: 4.6847 - acc: 0.0290 - ETA: 40s - loss: 4.6844 - acc: 0.0291 - ETA: 39s - loss: 4.6841 - acc: 0.0293 - ETA: 38s - loss: 4.6841 - acc: 0.0294 - ETA: 37s - loss: 4.6841 - acc: 0.0295 - ETA: 37s - loss: 4.6843 - acc: 0.0294 - ETA: 36s - loss: 4.6838 - acc: 0.0294 - ETA: 35s - loss: 4.6833 - acc: 0.0293 - ETA: 34s - loss: 4.6830 - acc: 0.0292 - ETA: 33s - loss: 4.6829 - acc: 0.0291 - ETA: 32s - loss: 4.6827 - acc: 0.0297 - ETA: 31s - loss: 4.6830 - acc: 0.0296 - ETA: 30s - loss: 4.6833 - acc: 0.0295 - ETA: 29s - loss: 4.6824 - acc: 0.0297 - ETA: 28s - loss: 4.6828 - acc: 0.0296 - ETA: 27s - loss: 4.6825 - acc: 0.0295 - ETA: 27s - loss: 4.6835 - acc: 0.0294 - ETA: 26s - loss: 4.6829 - acc: 0.0295 - ETA: 25s - loss: 4.6830 - acc: 0.0294 - ETA: 24s - loss: 4.6836 - acc: 0.0293 - ETA: 23s - loss: 4.6839 - acc: 0.0292 - ETA: 22s - loss: 4.6840 - acc: 0.0296 - ETA: 21s - loss: 4.6836 - acc: 0.0295 - ETA: 20s - loss: 4.6828 - acc: 0.0296 - ETA: 19s - loss: 4.6829 - acc: 0.0296 - ETA: 18s - loss: 4.6838 - acc: 0.0296 - ETA: 18s - loss: 4.6850 - acc: 0.0295 - ETA: 17s - loss: 4.6845 - acc: 0.0295 - ETA: 16s - loss: 4.6843 - acc: 0.0294 - ETA: 15s - loss: 4.6838 - acc: 0.0293 - ETA: 14s - loss: 4.6834 - acc: 0.0292 - ETA: 13s - loss: 4.6837 - acc: 0.0292 - ETA: 12s - loss: 4.6839 - acc: 0.0291 - ETA: 11s - loss: 4.6831 - acc: 0.0291 - ETA: 10s - loss: 4.6828 - acc: 0.0292 - ETA: 9s - loss: 4.6824 - acc: 0.0294  - ETA: 9s - loss: 4.6824 - acc: 0.0293 - ETA: 8s - loss: 4.6822 - acc: 0.0294 - ETA: 7s - loss: 4.6822 - acc: 0.0293 - ETA: 6s - loss: 4.6814 - acc: 0.0294 - ETA: 5s - loss: 4.6812 - acc: 0.0294 - ETA: 4s - loss: 4.6824 - acc: 0.0296 - ETA: 3s - loss: 4.6816 - acc: 0.0297 - ETA: 2s - loss: 4.6819 - acc: 0.0296 - ETA: 1s - loss: 4.6820 - acc: 0.0297 - ETA: 0s - loss: 4.6822 - acc: 0.0296Epoch 00007: val_loss improved from 4.74435 to 4.72558, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 317s - loss: 4.6839 - acc: 0.0295 - val_loss: 4.7256 - val_acc: 0.0228
    Epoch 9/10
    6660/6680 [============================>.] - ETA: 313s - loss: 4.6106 - acc: 0.0000e+00 - ETA: 308s - loss: 4.6969 - acc: 0.0000e+00 - ETA: 304s - loss: 4.7306 - acc: 0.0000e+00 - ETA: 306s - loss: 4.7640 - acc: 0.0000e+00 - ETA: 305s - loss: 4.7074 - acc: 0.0200     - ETA: 303s - loss: 4.6821 - acc: 0.0250 - ETA: 303s - loss: 4.6695 - acc: 0.0286 - ETA: 302s - loss: 4.6505 - acc: 0.0313 - ETA: 301s - loss: 4.6074 - acc: 0.0333 - ETA: 299s - loss: 4.6143 - acc: 0.0400 - ETA: 297s - loss: 4.6306 - acc: 0.0455 - ETA: 296s - loss: 4.6165 - acc: 0.0417 - ETA: 296s - loss: 4.6031 - acc: 0.0423 - ETA: 294s - loss: 4.6122 - acc: 0.0393 - ETA: 292s - loss: 4.6195 - acc: 0.0400 - ETA: 291s - loss: 4.6348 - acc: 0.0375 - ETA: 290s - loss: 4.6521 - acc: 0.0382 - ETA: 289s - loss: 4.6483 - acc: 0.0361 - ETA: 288s - loss: 4.6491 - acc: 0.0342 - ETA: 287s - loss: 4.6463 - acc: 0.0350 - ETA: 286s - loss: 4.6563 - acc: 0.0381 - ETA: 285s - loss: 4.6596 - acc: 0.0409 - ETA: 283s - loss: 4.6700 - acc: 0.0413 - ETA: 282s - loss: 4.6592 - acc: 0.0396 - ETA: 281s - loss: 4.6586 - acc: 0.0380 - ETA: 281s - loss: 4.6618 - acc: 0.0365 - ETA: 280s - loss: 4.6628 - acc: 0.0352 - ETA: 278s - loss: 4.6642 - acc: 0.0339 - ETA: 278s - loss: 4.6617 - acc: 0.0345 - ETA: 276s - loss: 4.6728 - acc: 0.0367 - ETA: 275s - loss: 4.6742 - acc: 0.0355 - ETA: 274s - loss: 4.6732 - acc: 0.0344 - ETA: 273s - loss: 4.6702 - acc: 0.0333 - ETA: 272s - loss: 4.6671 - acc: 0.0324 - ETA: 271s - loss: 4.6660 - acc: 0.0314 - ETA: 270s - loss: 4.6656 - acc: 0.0306 - ETA: 269s - loss: 4.6609 - acc: 0.0311 - ETA: 269s - loss: 4.6699 - acc: 0.0316 - ETA: 268s - loss: 4.6661 - acc: 0.0321 - ETA: 267s - loss: 4.6621 - acc: 0.0313 - ETA: 266s - loss: 4.6608 - acc: 0.0305 - ETA: 265s - loss: 4.6658 - acc: 0.0310 - ETA: 264s - loss: 4.6654 - acc: 0.0314 - ETA: 263s - loss: 4.6642 - acc: 0.0318 - ETA: 262s - loss: 4.6649 - acc: 0.0322 - ETA: 261s - loss: 4.6601 - acc: 0.0315 - ETA: 260s - loss: 4.6588 - acc: 0.0340 - ETA: 259s - loss: 4.6591 - acc: 0.0333 - ETA: 258s - loss: 4.6583 - acc: 0.0327 - ETA: 257s - loss: 4.6582 - acc: 0.0330 - ETA: 256s - loss: 4.6607 - acc: 0.0324 - ETA: 255s - loss: 4.6628 - acc: 0.0327 - ETA: 255s - loss: 4.6624 - acc: 0.0321 - ETA: 254s - loss: 4.6594 - acc: 0.0315 - ETA: 253s - loss: 4.6609 - acc: 0.0309 - ETA: 252s - loss: 4.6590 - acc: 0.0313 - ETA: 251s - loss: 4.6540 - acc: 0.0316 - ETA: 250s - loss: 4.6583 - acc: 0.0319 - ETA: 249s - loss: 4.6587 - acc: 0.0314 - ETA: 248s - loss: 4.6551 - acc: 0.0317 - ETA: 247s - loss: 4.6558 - acc: 0.0311 - ETA: 246s - loss: 4.6513 - acc: 0.0315 - ETA: 245s - loss: 4.6518 - acc: 0.0310 - ETA: 245s - loss: 4.6476 - acc: 0.0320 - ETA: 244s - loss: 4.6477 - acc: 0.0315 - ETA: 243s - loss: 4.6477 - acc: 0.0318 - ETA: 242s - loss: 4.6469 - acc: 0.0313 - ETA: 241s - loss: 4.6466 - acc: 0.0316 - ETA: 240s - loss: 4.6464 - acc: 0.0319 - ETA: 240s - loss: 4.6480 - acc: 0.0321 - ETA: 239s - loss: 4.6459 - acc: 0.0317 - ETA: 238s - loss: 4.6432 - acc: 0.0319 - ETA: 237s - loss: 4.6465 - acc: 0.0315 - ETA: 236s - loss: 4.6462 - acc: 0.0318 - ETA: 235s - loss: 4.6472 - acc: 0.0313 - ETA: 234s - loss: 4.6482 - acc: 0.0309 - ETA: 233s - loss: 4.6453 - acc: 0.0305 - ETA: 232s - loss: 4.6478 - acc: 0.0308 - ETA: 231s - loss: 4.6451 - acc: 0.0310 - ETA: 230s - loss: 4.6432 - acc: 0.0306 - ETA: 229s - loss: 4.6418 - acc: 0.0309 - ETA: 228s - loss: 4.6455 - acc: 0.0305 - ETA: 227s - loss: 4.6451 - acc: 0.0307 - ETA: 226s - loss: 4.6440 - acc: 0.0315 - ETA: 226s - loss: 4.6428 - acc: 0.0312 - ETA: 225s - loss: 4.6423 - acc: 0.0308 - ETA: 224s - loss: 4.6429 - acc: 0.0310 - ETA: 223s - loss: 4.6468 - acc: 0.0313 - ETA: 222s - loss: 4.6446 - acc: 0.0309 - ETA: 221s - loss: 4.6440 - acc: 0.0311 - ETA: 220s - loss: 4.6430 - acc: 0.0319 - ETA: 219s - loss: 4.6427 - acc: 0.0315 - ETA: 218s - loss: 4.6431 - acc: 0.0312 - ETA: 217s - loss: 4.6432 - acc: 0.0319 - ETA: 216s - loss: 4.6448 - acc: 0.0316 - ETA: 215s - loss: 4.6433 - acc: 0.0313 - ETA: 214s - loss: 4.6445 - acc: 0.0309 - ETA: 213s - loss: 4.6444 - acc: 0.0311 - ETA: 212s - loss: 4.6434 - acc: 0.0318 - ETA: 211s - loss: 4.6419 - acc: 0.0315 - ETA: 210s - loss: 4.6414 - acc: 0.0317 - ETA: 209s - loss: 4.6425 - acc: 0.0319 - ETA: 208s - loss: 4.6426 - acc: 0.0325 - ETA: 207s - loss: 4.6419 - acc: 0.0327 - ETA: 207s - loss: 4.6427 - acc: 0.0324 - ETA: 206s - loss: 4.6398 - acc: 0.0330 - ETA: 205s - loss: 4.6393 - acc: 0.0327 - ETA: 204s - loss: 4.6413 - acc: 0.0329 - ETA: 203s - loss: 4.6410 - acc: 0.0330 - ETA: 202s - loss: 4.6403 - acc: 0.0327 - ETA: 201s - loss: 4.6398 - acc: 0.0324 - ETA: 200s - loss: 4.6404 - acc: 0.0326 - ETA: 199s - loss: 4.6413 - acc: 0.0323 - ETA: 198s - loss: 4.6432 - acc: 0.0320 - ETA: 197s - loss: 4.6457 - acc: 0.0322 - ETA: 196s - loss: 4.6450 - acc: 0.0319 - ETA: 195s - loss: 4.6442 - acc: 0.0325 - ETA: 194s - loss: 4.6447 - acc: 0.0326 - ETA: 194s - loss: 4.6469 - acc: 0.0324 - ETA: 193s - loss: 4.6466 - acc: 0.0325 - ETA: 192s - loss: 4.6458 - acc: 0.0335 - ETA: 191s - loss: 4.6467 - acc: 0.0336 - ETA: 190s - loss: 4.6477 - acc: 0.0333 - ETA: 189s - loss: 4.6495 - acc: 0.0331 - ETA: 188s - loss: 4.6495 - acc: 0.0328 - ETA: 187s - loss: 4.6493 - acc: 0.0325 - ETA: 186s - loss: 4.6493 - acc: 0.0323 - ETA: 185s - loss: 4.6503 - acc: 0.0320 - ETA: 184s - loss: 4.6509 - acc: 0.0318 - ETA: 183s - loss: 4.6512 - acc: 0.0315 - ETA: 182s - loss: 4.6507 - acc: 0.0313 - ETA: 181s - loss: 4.6500 - acc: 0.0314 - ETA: 181s - loss: 4.6497 - acc: 0.0316 - ETA: 180s - loss: 4.6497 - acc: 0.0321 - ETA: 179s - loss: 4.6496 - acc: 0.0319 - ETA: 178s - loss: 4.6491 - acc: 0.0316 - ETA: 177s - loss: 4.6495 - acc: 0.0314 - ETA: 176s - loss: 4.6511 - acc: 0.0312 - ETA: 175s - loss: 4.6516 - acc: 0.0313 - ETA: 174s - loss: 4.6504 - acc: 0.0311 - ETA: 173s - loss: 4.6512 - acc: 0.0312 - ETA: 172s - loss: 4.6543 - acc: 0.0313 - ETA: 171s - loss: 4.6572 - acc: 0.0311 - ETA: 171s - loss: 4.6585 - acc: 0.0309 - ETA: 170s - loss: 4.6577 - acc: 0.0310 - ETA: 169s - loss: 4.6580 - acc: 0.0308 - ETA: 168s - loss: 4.6581 - acc: 0.0306 - ETA: 167s - loss: 4.6574 - acc: 0.0304 - ETA: 166s - loss: 4.6589 - acc: 0.0305 - ETA: 165s - loss: 4.6585 - acc: 0.0303 - ETA: 164s - loss: 4.6587 - acc: 0.0301 - ETA: 163s - loss: 4.6597 - acc: 0.0299 - ETA: 162s - loss: 4.6574 - acc: 0.0307 - ETA: 161s - loss: 4.6572 - acc: 0.0308 - ETA: 161s - loss: 4.6563 - acc: 0.0306 - ETA: 160s - loss: 4.6567 - acc: 0.0311 - ETA: 159s - loss: 4.6565 - acc: 0.0309 - ETA: 158s - loss: 4.6567 - acc: 0.0310 - ETA: 157s - loss: 4.6581 - acc: 0.0308 - ETA: 156s - loss: 4.6586 - acc: 0.0306 - ETA: 155s - loss: 4.6591 - acc: 0.0304 - ETA: 154s - loss: 4.6584 - acc: 0.0302 - ETA: 153s - loss: 4.6591 - acc: 0.0301 - ETA: 152s - loss: 4.6588 - acc: 0.0299 - ETA: 151s - loss: 4.6574 - acc: 0.0306 - ETA: 151s - loss: 4.6583 - acc: 0.0307 - ETA: 150s - loss: 4.6586 - acc: 0.0311 - ETA: 149s - loss: 4.6581 - acc: 0.0315 - ETA: 148s - loss: 4.6587 - acc: 0.0314 - ETA: 147s - loss: 4.6585 - acc: 0.0312 - ETA: 146s - loss: 4.6587 - acc: 0.0313 - ETA: 145s - loss: 4.6579 - acc: 0.0317 - ETA: 144s - loss: 4.6585 - acc: 0.0315 - ETA: 143s - loss: 4.6580 - acc: 0.0316 - ETA: 142s - loss: 4.6567 - acc: 0.0317 - ETA: 141s - loss: 4.6578 - acc: 0.0315 - ETA: 141s - loss: 4.6574 - acc: 0.0314 - ETA: 140s - loss: 4.6571 - acc: 0.0312 - ETA: 139s - loss: 4.6567 - acc: 0.0316 - ETA: 138s - loss: 4.6561 - acc: 0.0317 - ETA: 137s - loss: 4.6560 - acc: 0.0320 - ETA: 136s - loss: 4.6560 - acc: 0.0321 - ETA: 135s - loss: 4.6572 - acc: 0.0320 - ETA: 134s - loss: 4.6568 - acc: 0.0318 - ETA: 133s - loss: 4.6566 - acc: 0.0316 - ETA: 132s - loss: 4.6570 - acc: 0.0317 - ETA: 132s - loss: 4.6566 - acc: 0.0316 - ETA: 131s - loss: 4.6563 - acc: 0.0314 - ETA: 130s - loss: 4.6568 - acc: 0.0315 - ETA: 129s - loss: 4.6569 - acc: 0.0316 - ETA: 128s - loss: 4.6572 - acc: 0.0314 - ETA: 127s - loss: 4.6577 - acc: 0.0313 - ETA: 126s - loss: 4.6579 - acc: 0.0311 - ETA: 125s - loss: 4.6569 - acc: 0.0309 - ETA: 124s - loss: 4.6572 - acc: 0.0313 - ETA: 123s - loss: 4.6570 - acc: 0.0311 - ETA: 122s - loss: 4.6563 - acc: 0.0312 - ETA: 122s - loss: 4.6557 - acc: 0.0311 - ETA: 121s - loss: 4.6559 - acc: 0.0309 - ETA: 120s - loss: 4.6569 - acc: 0.0308 - ETA: 119s - loss: 4.6555 - acc: 0.0308 - ETA: 118s - loss: 4.6552 - acc: 0.0309 - ETA: 117s - loss: 4.6554 - acc: 0.0313 - ETA: 116s - loss: 4.6548 - acc: 0.0316 - ETA: 115s - loss: 4.6549 - acc: 0.0315 - ETA: 114s - loss: 4.6543 - acc: 0.0316 - ETA: 113s - loss: 4.6541 - acc: 0.0324 - ETA: 113s - loss: 4.6544 - acc: 0.0322 - ETA: 112s - loss: 4.6544 - acc: 0.0321 - ETA: 111s - loss: 4.6548 - acc: 0.0321 - ETA: 110s - loss: 4.6560 - acc: 0.0325 - ETA: 109s - loss: 4.6561 - acc: 0.0323 - ETA: 108s - loss: 4.6560 - acc: 0.0322 - ETA: 107s - loss: 4.6554 - acc: 0.0322 - ETA: 106s - loss: 4.6555 - acc: 0.0326 - ETA: 105s - loss: 4.6554 - acc: 0.0326 - ETA: 104s - loss: 4.6559 - acc: 0.0327 - ETA: 104s - loss: 4.6554 - acc: 0.0326 - ETA: 103s - loss: 4.6563 - acc: 0.0326 - ETA: 102s - loss: 4.6554 - acc: 0.0327 - ETA: 101s - loss: 4.6569 - acc: 0.0326 - ETA: 100s - loss: 4.6566 - acc: 0.0329 - ETA: 99s - loss: 4.6555 - acc: 0.0327  - ETA: 98s - loss: 4.6561 - acc: 0.0326 - ETA: 97s - loss: 4.6559 - acc: 0.0327 - ETA: 96s - loss: 4.6561 - acc: 0.0327 - ETA: 95s - loss: 4.6557 - acc: 0.0328 - ETA: 95s - loss: 4.6562 - acc: 0.0329 - ETA: 94s - loss: 4.6554 - acc: 0.0330 - ETA: 93s - loss: 4.6545 - acc: 0.0328 - ETA: 92s - loss: 4.6552 - acc: 0.0329 - ETA: 91s - loss: 4.6559 - acc: 0.0328 - ETA: 90s - loss: 4.6563 - acc: 0.0326 - ETA: 89s - loss: 4.6564 - acc: 0.0329 - ETA: 88s - loss: 4.6579 - acc: 0.0328 - ETA: 87s - loss: 4.6578 - acc: 0.0331 - ETA: 87s - loss: 4.6576 - acc: 0.0331 - ETA: 86s - loss: 4.6574 - acc: 0.0332 - ETA: 85s - loss: 4.6569 - acc: 0.0335 - ETA: 84s - loss: 4.6580 - acc: 0.0333 - ETA: 83s - loss: 4.6577 - acc: 0.0338 - ETA: 82s - loss: 4.6572 - acc: 0.0341 - ETA: 81s - loss: 4.6571 - acc: 0.0340 - ETA: 80s - loss: 4.6568 - acc: 0.0342 - ETA: 79s - loss: 4.6572 - acc: 0.0343 - ETA: 78s - loss: 4.6569 - acc: 0.0343 - ETA: 78s - loss: 4.6570 - acc: 0.0342 - ETA: 77s - loss: 4.6563 - acc: 0.0345 - ETA: 76s - loss: 4.6564 - acc: 0.0343 - ETA: 75s - loss: 4.6572 - acc: 0.0344 - ETA: 74s - loss: 4.6568 - acc: 0.0343 - ETA: 73s - loss: 4.6563 - acc: 0.0341 - ETA: 72s - loss: 4.6565 - acc: 0.0340 - ETA: 71s - loss: 4.6580 - acc: 0.0339 - ETA: 70s - loss: 4.6574 - acc: 0.0337 - ETA: 70s - loss: 4.6571 - acc: 0.0338 - ETA: 69s - loss: 4.6568 - acc: 0.0339 - ETA: 68s - loss: 4.6569 - acc: 0.0337 - ETA: 67s - loss: 4.6569 - acc: 0.0336 - ETA: 66s - loss: 4.6581 - acc: 0.0335 - ETA: 65s - loss: 4.6583 - acc: 0.0333 - ETA: 64s - loss: 4.6594 - acc: 0.0334 - ETA: 63s - loss: 4.6591 - acc: 0.0337 - ETA: 62s - loss: 4.6586 - acc: 0.0337 - ETA: 61s - loss: 4.6591 - acc: 0.0338 - ETA: 61s - loss: 4.6590 - acc: 0.0336 - ETA: 60s - loss: 4.6589 - acc: 0.0337 - ETA: 59s - loss: 4.6587 - acc: 0.0338 - ETA: 58s - loss: 4.6585 - acc: 0.0338 - ETA: 57s - loss: 4.6595 - acc: 0.0337 - ETA: 56s - loss: 4.6593 - acc: 0.0336 - ETA: 55s - loss: 4.6594 - acc: 0.0336 - ETA: 54s - loss: 4.6591 - acc: 0.0337 - ETA: 53s - loss: 4.6590 - acc: 0.0336 - ETA: 52s - loss: 4.6589 - acc: 0.0336 - ETA: 52s - loss: 4.6589 - acc: 0.0335 - ETA: 51s - loss: 4.6587 - acc: 0.0334 - ETA: 50s - loss: 4.6588 - acc: 0.0333 - ETA: 49s - loss: 4.6584 - acc: 0.0335 - ETA: 48s - loss: 4.6584 - acc: 0.0336 - ETA: 47s - loss: 4.6587 - acc: 0.0338 - ETA: 46s - loss: 4.6588 - acc: 0.0337 - ETA: 45s - loss: 4.6592 - acc: 0.0336 - ETA: 44s - loss: 4.6593 - acc: 0.0336 - ETA: 43s - loss: 4.6595 - acc: 0.0335 - ETA: 43s - loss: 4.6599 - acc: 0.0334 - ETA: 42s - loss: 4.6596 - acc: 0.0333 - ETA: 41s - loss: 4.6588 - acc: 0.0335 - ETA: 40s - loss: 4.6588 - acc: 0.0334 - ETA: 39s - loss: 4.6585 - acc: 0.0334 - ETA: 38s - loss: 4.6586 - acc: 0.0337 - ETA: 37s - loss: 4.6578 - acc: 0.0336 - ETA: 36s - loss: 4.6581 - acc: 0.0336 - ETA: 35s - loss: 4.6572 - acc: 0.0335 - ETA: 34s - loss: 4.6570 - acc: 0.0337 - ETA: 34s - loss: 4.6571 - acc: 0.0336 - ETA: 33s - loss: 4.6565 - acc: 0.0337 - ETA: 32s - loss: 4.6570 - acc: 0.0337 - ETA: 31s - loss: 4.6575 - acc: 0.0336 - ETA: 30s - loss: 4.6579 - acc: 0.0335 - ETA: 29s - loss: 4.6586 - acc: 0.0334 - ETA: 28s - loss: 4.6585 - acc: 0.0334 - ETA: 27s - loss: 4.6579 - acc: 0.0335 - ETA: 26s - loss: 4.6576 - acc: 0.0334 - ETA: 26s - loss: 4.6581 - acc: 0.0333 - ETA: 25s - loss: 4.6580 - acc: 0.0332 - ETA: 24s - loss: 4.6577 - acc: 0.0334 - ETA: 23s - loss: 4.6571 - acc: 0.0333 - ETA: 22s - loss: 4.6570 - acc: 0.0335 - ETA: 21s - loss: 4.6566 - acc: 0.0334 - ETA: 20s - loss: 4.6565 - acc: 0.0336 - ETA: 19s - loss: 4.6562 - acc: 0.0335 - ETA: 18s - loss: 4.6555 - acc: 0.0337 - ETA: 17s - loss: 4.6562 - acc: 0.0336 - ETA: 17s - loss: 4.6559 - acc: 0.0335 - ETA: 16s - loss: 4.6560 - acc: 0.0334 - ETA: 15s - loss: 4.6557 - acc: 0.0333 - ETA: 14s - loss: 4.6569 - acc: 0.0332 - ETA: 13s - loss: 4.6571 - acc: 0.0331 - ETA: 12s - loss: 4.6572 - acc: 0.0331 - ETA: 11s - loss: 4.6570 - acc: 0.0333 - ETA: 10s - loss: 4.6568 - acc: 0.0334 - ETA: 9s - loss: 4.6568 - acc: 0.0333  - ETA: 8s - loss: 4.6569 - acc: 0.0333 - ETA: 8s - loss: 4.6560 - acc: 0.0335 - ETA: 7s - loss: 4.6563 - acc: 0.0336 - ETA: 6s - loss: 4.6560 - acc: 0.0335 - ETA: 5s - loss: 4.6563 - acc: 0.0335 - ETA: 4s - loss: 4.6563 - acc: 0.0334 - ETA: 3s - loss: 4.6559 - acc: 0.0336 - ETA: 2s - loss: 4.6559 - acc: 0.0337 - ETA: 1s - loss: 4.6559 - acc: 0.0337 - ETA: 0s - loss: 4.6555 - acc: 0.0338Epoch 00008: val_loss improved from 4.72558 to 4.69424, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 315s - loss: 4.6552 - acc: 0.0338 - val_loss: 4.6942 - val_acc: 0.0287
    Epoch 10/10
    6660/6680 [============================>.] - ETA: 291s - loss: 4.4873 - acc: 0.0000e+00 - ETA: 296s - loss: 4.5930 - acc: 0.0250     - ETA: 297s - loss: 4.5695 - acc: 0.0333 - ETA: 297s - loss: 4.5928 - acc: 0.0375 - ETA: 297s - loss: 4.6122 - acc: 0.0300 - ETA: 297s - loss: 4.6044 - acc: 0.0333 - ETA: 295s - loss: 4.6896 - acc: 0.0357 - ETA: 293s - loss: 4.6965 - acc: 0.0375 - ETA: 291s - loss: 4.6867 - acc: 0.0333 - ETA: 291s - loss: 4.6630 - acc: 0.0400 - ETA: 290s - loss: 4.6798 - acc: 0.0364 - ETA: 289s - loss: 4.6636 - acc: 0.0417 - ETA: 288s - loss: 4.6496 - acc: 0.0423 - ETA: 288s - loss: 4.6581 - acc: 0.0393 - ETA: 287s - loss: 4.6546 - acc: 0.0400 - ETA: 286s - loss: 4.6509 - acc: 0.0406 - ETA: 285s - loss: 4.6486 - acc: 0.0382 - ETA: 284s - loss: 4.6469 - acc: 0.0389 - ETA: 284s - loss: 4.6533 - acc: 0.0368 - ETA: 283s - loss: 4.6585 - acc: 0.0400 - ETA: 282s - loss: 4.6487 - acc: 0.0429 - ETA: 281s - loss: 4.6544 - acc: 0.0409 - ETA: 280s - loss: 4.6488 - acc: 0.0391 - ETA: 279s - loss: 4.6392 - acc: 0.0375 - ETA: 278s - loss: 4.6478 - acc: 0.0360 - ETA: 277s - loss: 4.6424 - acc: 0.0365 - ETA: 276s - loss: 4.6451 - acc: 0.0352 - ETA: 276s - loss: 4.6467 - acc: 0.0375 - ETA: 275s - loss: 4.6403 - acc: 0.0397 - ETA: 275s - loss: 4.6404 - acc: 0.0383 - ETA: 274s - loss: 4.6417 - acc: 0.0387 - ETA: 273s - loss: 4.6500 - acc: 0.0375 - ETA: 272s - loss: 4.6508 - acc: 0.0364 - ETA: 271s - loss: 4.6519 - acc: 0.0353 - ETA: 270s - loss: 4.6510 - acc: 0.0371 - ETA: 269s - loss: 4.6462 - acc: 0.0375 - ETA: 268s - loss: 4.6428 - acc: 0.0365 - ETA: 267s - loss: 4.6326 - acc: 0.0382 - ETA: 267s - loss: 4.6283 - acc: 0.0372 - ETA: 266s - loss: 4.6269 - acc: 0.0400 - ETA: 265s - loss: 4.6284 - acc: 0.0390 - ETA: 263s - loss: 4.6305 - acc: 0.0381 - ETA: 262s - loss: 4.6299 - acc: 0.0372 - ETA: 261s - loss: 4.6223 - acc: 0.0375 - ETA: 260s - loss: 4.6217 - acc: 0.0378 - ETA: 259s - loss: 4.6199 - acc: 0.0380 - ETA: 258s - loss: 4.6142 - acc: 0.0383 - ETA: 257s - loss: 4.6177 - acc: 0.0385 - ETA: 256s - loss: 4.6187 - acc: 0.0378 - ETA: 256s - loss: 4.6205 - acc: 0.0370 - ETA: 255s - loss: 4.6171 - acc: 0.0373 - ETA: 254s - loss: 4.6175 - acc: 0.0375 - ETA: 253s - loss: 4.6200 - acc: 0.0368 - ETA: 252s - loss: 4.6182 - acc: 0.0361 - ETA: 251s - loss: 4.6170 - acc: 0.0355 - ETA: 250s - loss: 4.6155 - acc: 0.0357 - ETA: 249s - loss: 4.6165 - acc: 0.0351 - ETA: 248s - loss: 4.6222 - acc: 0.0345 - ETA: 247s - loss: 4.6173 - acc: 0.0356 - ETA: 246s - loss: 4.6165 - acc: 0.0350 - ETA: 246s - loss: 4.6180 - acc: 0.0352 - ETA: 245s - loss: 4.6204 - acc: 0.0347 - ETA: 244s - loss: 4.6201 - acc: 0.0341 - ETA: 243s - loss: 4.6220 - acc: 0.0344 - ETA: 242s - loss: 4.6216 - acc: 0.0346 - ETA: 241s - loss: 4.6197 - acc: 0.0341 - ETA: 240s - loss: 4.6231 - acc: 0.0336 - ETA: 239s - loss: 4.6238 - acc: 0.0346 - ETA: 238s - loss: 4.6193 - acc: 0.0348 - ETA: 237s - loss: 4.6151 - acc: 0.0343 - ETA: 236s - loss: 4.6123 - acc: 0.0345 - ETA: 235s - loss: 4.6116 - acc: 0.0347 - ETA: 234s - loss: 4.6098 - acc: 0.0349 - ETA: 233s - loss: 4.6093 - acc: 0.0345 - ETA: 233s - loss: 4.6078 - acc: 0.0347 - ETA: 231s - loss: 4.6062 - acc: 0.0355 - ETA: 230s - loss: 4.6046 - acc: 0.0357 - ETA: 230s - loss: 4.6085 - acc: 0.0353 - ETA: 229s - loss: 4.6087 - acc: 0.0348 - ETA: 228s - loss: 4.6093 - acc: 0.0344 - ETA: 227s - loss: 4.6114 - acc: 0.0340 - ETA: 226s - loss: 4.6098 - acc: 0.0335 - ETA: 225s - loss: 4.6089 - acc: 0.0331 - ETA: 224s - loss: 4.6122 - acc: 0.0327 - ETA: 223s - loss: 4.6138 - acc: 0.0329 - ETA: 222s - loss: 4.6124 - acc: 0.0337 - ETA: 221s - loss: 4.6134 - acc: 0.0339 - ETA: 220s - loss: 4.6144 - acc: 0.0335 - ETA: 220s - loss: 4.6160 - acc: 0.0337 - ETA: 219s - loss: 4.6150 - acc: 0.0344 - ETA: 218s - loss: 4.6156 - acc: 0.0341 - ETA: 217s - loss: 4.6139 - acc: 0.0337 - ETA: 216s - loss: 4.6134 - acc: 0.0339 - ETA: 215s - loss: 4.6144 - acc: 0.0340 - ETA: 214s - loss: 4.6159 - acc: 0.0342 - ETA: 213s - loss: 4.6182 - acc: 0.0344 - ETA: 213s - loss: 4.6204 - acc: 0.0340 - ETA: 212s - loss: 4.6192 - acc: 0.0337 - ETA: 211s - loss: 4.6215 - acc: 0.0333 - ETA: 210s - loss: 4.6213 - acc: 0.0330 - ETA: 209s - loss: 4.6212 - acc: 0.0327 - ETA: 208s - loss: 4.6209 - acc: 0.0328 - ETA: 207s - loss: 4.6208 - acc: 0.0330 - ETA: 206s - loss: 4.6200 - acc: 0.0332 - ETA: 205s - loss: 4.6228 - acc: 0.0329 - ETA: 204s - loss: 4.6217 - acc: 0.0325 - ETA: 203s - loss: 4.6210 - acc: 0.0322 - ETA: 203s - loss: 4.6194 - acc: 0.0324 - ETA: 202s - loss: 4.6176 - acc: 0.0330 - ETA: 201s - loss: 4.6159 - acc: 0.0336 - ETA: 200s - loss: 4.6138 - acc: 0.0347 - ETA: 199s - loss: 4.6150 - acc: 0.0344 - ETA: 198s - loss: 4.6157 - acc: 0.0345 - ETA: 197s - loss: 4.6155 - acc: 0.0346 - ETA: 196s - loss: 4.6176 - acc: 0.0348 - ETA: 195s - loss: 4.6178 - acc: 0.0345 - ETA: 194s - loss: 4.6182 - acc: 0.0350 - ETA: 193s - loss: 4.6181 - acc: 0.0356 - ETA: 192s - loss: 4.6179 - acc: 0.0357 - ETA: 192s - loss: 4.6172 - acc: 0.0358 - ETA: 191s - loss: 4.6174 - acc: 0.0364 - ETA: 190s - loss: 4.6159 - acc: 0.0365 - ETA: 189s - loss: 4.6160 - acc: 0.0362 - ETA: 188s - loss: 4.6163 - acc: 0.0367 - ETA: 187s - loss: 4.6156 - acc: 0.0372 - ETA: 186s - loss: 4.6175 - acc: 0.0377 - ETA: 185s - loss: 4.6168 - acc: 0.0382 - ETA: 184s - loss: 4.6157 - acc: 0.0387 - ETA: 183s - loss: 4.6174 - acc: 0.0384 - ETA: 182s - loss: 4.6168 - acc: 0.0381 - ETA: 182s - loss: 4.6144 - acc: 0.0385 - ETA: 181s - loss: 4.6139 - acc: 0.0386 - ETA: 180s - loss: 4.6153 - acc: 0.0383 - ETA: 179s - loss: 4.6156 - acc: 0.0381 - ETA: 178s - loss: 4.6147 - acc: 0.0378 - ETA: 177s - loss: 4.6147 - acc: 0.0382 - ETA: 176s - loss: 4.6154 - acc: 0.0383 - ETA: 175s - loss: 4.6148 - acc: 0.0384 - ETA: 174s - loss: 4.6163 - acc: 0.0381 - ETA: 174s - loss: 4.6152 - acc: 0.0379 - ETA: 173s - loss: 4.6161 - acc: 0.0376 - ETA: 172s - loss: 4.6158 - acc: 0.0377 - ETA: 171s - loss: 4.6167 - acc: 0.0378 - ETA: 170s - loss: 4.6180 - acc: 0.0378 - ETA: 169s - loss: 4.6158 - acc: 0.0383 - ETA: 168s - loss: 4.6168 - acc: 0.0380 - ETA: 167s - loss: 4.6180 - acc: 0.0378 - ETA: 166s - loss: 4.6194 - acc: 0.0378 - ETA: 165s - loss: 4.6189 - acc: 0.0376 - ETA: 164s - loss: 4.6181 - acc: 0.0377 - ETA: 164s - loss: 4.6173 - acc: 0.0377 - ETA: 163s - loss: 4.6176 - acc: 0.0375 - ETA: 162s - loss: 4.6163 - acc: 0.0379 - ETA: 161s - loss: 4.6164 - acc: 0.0377 - ETA: 160s - loss: 4.6161 - acc: 0.0381 - ETA: 159s - loss: 4.6180 - acc: 0.0378 - ETA: 158s - loss: 4.6170 - acc: 0.0379 - ETA: 157s - loss: 4.6155 - acc: 0.0377 - ETA: 156s - loss: 4.6158 - acc: 0.0374 - ETA: 155s - loss: 4.6147 - acc: 0.0372 - ETA: 155s - loss: 4.6150 - acc: 0.0370 - ETA: 154s - loss: 4.6150 - acc: 0.0370 - ETA: 153s - loss: 4.6161 - acc: 0.0368 - ETA: 152s - loss: 4.6169 - acc: 0.0369 - ETA: 151s - loss: 4.6153 - acc: 0.0373 - ETA: 150s - loss: 4.6157 - acc: 0.0370 - ETA: 149s - loss: 4.6171 - acc: 0.0374 - ETA: 148s - loss: 4.6174 - acc: 0.0372 - ETA: 147s - loss: 4.6180 - acc: 0.0376 - ETA: 146s - loss: 4.6187 - acc: 0.0374 - ETA: 145s - loss: 4.6204 - acc: 0.0374 - ETA: 144s - loss: 4.6206 - acc: 0.0372 - ETA: 144s - loss: 4.6204 - acc: 0.0370 - ETA: 143s - loss: 4.6214 - acc: 0.0368 - ETA: 142s - loss: 4.6222 - acc: 0.0369 - ETA: 141s - loss: 4.6224 - acc: 0.0369 - ETA: 140s - loss: 4.6230 - acc: 0.0370 - ETA: 139s - loss: 4.6228 - acc: 0.0371 - ETA: 138s - loss: 4.6232 - acc: 0.0369 - ETA: 137s - loss: 4.6229 - acc: 0.0367 - ETA: 136s - loss: 4.6227 - acc: 0.0365 - ETA: 136s - loss: 4.6227 - acc: 0.0365 - ETA: 135s - loss: 4.6225 - acc: 0.0369 - ETA: 134s - loss: 4.6219 - acc: 0.0367 - ETA: 133s - loss: 4.6230 - acc: 0.0365 - ETA: 132s - loss: 4.6216 - acc: 0.0366 - ETA: 131s - loss: 4.6225 - acc: 0.0366 - ETA: 130s - loss: 4.6236 - acc: 0.0370 - ETA: 129s - loss: 4.6237 - acc: 0.0368 - ETA: 128s - loss: 4.6249 - acc: 0.0366 - ETA: 127s - loss: 4.6250 - acc: 0.0364 - ETA: 127s - loss: 4.6256 - acc: 0.0362 - ETA: 126s - loss: 4.6262 - acc: 0.0360 - ETA: 125s - loss: 4.6261 - acc: 0.0366 - ETA: 124s - loss: 4.6267 - acc: 0.0367 - ETA: 123s - loss: 4.6261 - acc: 0.0367 - ETA: 122s - loss: 4.6267 - acc: 0.0365 - ETA: 121s - loss: 4.6260 - acc: 0.0369 - ETA: 120s - loss: 4.6270 - acc: 0.0367 - ETA: 120s - loss: 4.6263 - acc: 0.0368 - ETA: 119s - loss: 4.6264 - acc: 0.0368 - ETA: 118s - loss: 4.6271 - acc: 0.0366 - ETA: 117s - loss: 4.6263 - acc: 0.0372 - ETA: 116s - loss: 4.6271 - acc: 0.0370 - ETA: 115s - loss: 4.6270 - acc: 0.0368 - ETA: 114s - loss: 4.6263 - acc: 0.0369 - ETA: 113s - loss: 4.6269 - acc: 0.0367 - ETA: 112s - loss: 4.6256 - acc: 0.0368 - ETA: 112s - loss: 4.6258 - acc: 0.0366 - ETA: 111s - loss: 4.6262 - acc: 0.0369 - ETA: 110s - loss: 4.6265 - acc: 0.0370 - ETA: 109s - loss: 4.6267 - acc: 0.0370 - ETA: 108s - loss: 4.6260 - acc: 0.0371 - ETA: 107s - loss: 4.6246 - acc: 0.0371 - ETA: 106s - loss: 4.6259 - acc: 0.0370 - ETA: 105s - loss: 4.6264 - acc: 0.0368 - ETA: 104s - loss: 4.6267 - acc: 0.0369 - ETA: 104s - loss: 4.6268 - acc: 0.0369 - ETA: 103s - loss: 4.6277 - acc: 0.0370 - ETA: 102s - loss: 4.6270 - acc: 0.0370 - ETA: 101s - loss: 4.6275 - acc: 0.0369 - ETA: 100s - loss: 4.6271 - acc: 0.0367 - ETA: 99s - loss: 4.6265 - acc: 0.0368  - ETA: 98s - loss: 4.6264 - acc: 0.0373 - ETA: 97s - loss: 4.6269 - acc: 0.0371 - ETA: 96s - loss: 4.6260 - acc: 0.0372 - ETA: 96s - loss: 4.6248 - acc: 0.0374 - ETA: 95s - loss: 4.6248 - acc: 0.0375 - ETA: 94s - loss: 4.6256 - acc: 0.0378 - ETA: 93s - loss: 4.6245 - acc: 0.0378 - ETA: 92s - loss: 4.6237 - acc: 0.0377 - ETA: 91s - loss: 4.6233 - acc: 0.0375 - ETA: 90s - loss: 4.6237 - acc: 0.0376 - ETA: 89s - loss: 4.6240 - acc: 0.0376 - ETA: 88s - loss: 4.6252 - acc: 0.0374 - ETA: 87s - loss: 4.6258 - acc: 0.0377 - ETA: 87s - loss: 4.6260 - acc: 0.0378 - ETA: 86s - loss: 4.6257 - acc: 0.0376 - ETA: 85s - loss: 4.6253 - acc: 0.0377 - ETA: 84s - loss: 4.6269 - acc: 0.0375 - ETA: 83s - loss: 4.6267 - acc: 0.0378 - ETA: 82s - loss: 4.6269 - acc: 0.0378 - ETA: 81s - loss: 4.6267 - acc: 0.0377 - ETA: 80s - loss: 4.6272 - acc: 0.0379 - ETA: 79s - loss: 4.6268 - acc: 0.0378 - ETA: 78s - loss: 4.6271 - acc: 0.0378 - ETA: 78s - loss: 4.6270 - acc: 0.0377 - ETA: 77s - loss: 4.6276 - acc: 0.0375 - ETA: 76s - loss: 4.6278 - acc: 0.0373 - ETA: 75s - loss: 4.6286 - acc: 0.0374 - ETA: 74s - loss: 4.6279 - acc: 0.0373 - ETA: 73s - loss: 4.6278 - acc: 0.0373 - ETA: 72s - loss: 4.6276 - acc: 0.0374 - ETA: 71s - loss: 4.6264 - acc: 0.0372 - ETA: 70s - loss: 4.6266 - acc: 0.0371 - ETA: 70s - loss: 4.6259 - acc: 0.0369 - ETA: 69s - loss: 4.6254 - acc: 0.0370 - ETA: 68s - loss: 4.6244 - acc: 0.0368 - ETA: 67s - loss: 4.6239 - acc: 0.0369 - ETA: 66s - loss: 4.6235 - acc: 0.0369 - ETA: 65s - loss: 4.6234 - acc: 0.0368 - ETA: 64s - loss: 4.6237 - acc: 0.0366 - ETA: 63s - loss: 4.6235 - acc: 0.0367 - ETA: 62s - loss: 4.6237 - acc: 0.0366 - ETA: 61s - loss: 4.6240 - acc: 0.0364 - ETA: 61s - loss: 4.6241 - acc: 0.0363 - ETA: 60s - loss: 4.6234 - acc: 0.0363 - ETA: 59s - loss: 4.6225 - acc: 0.0364 - ETA: 58s - loss: 4.6217 - acc: 0.0364 - ETA: 57s - loss: 4.6210 - acc: 0.0365 - ETA: 56s - loss: 4.6218 - acc: 0.0363 - ETA: 55s - loss: 4.6214 - acc: 0.0364 - ETA: 54s - loss: 4.6215 - acc: 0.0364 - ETA: 53s - loss: 4.6210 - acc: 0.0363 - ETA: 52s - loss: 4.6205 - acc: 0.0364 - ETA: 52s - loss: 4.6197 - acc: 0.0364 - ETA: 51s - loss: 4.6196 - acc: 0.0365 - ETA: 50s - loss: 4.6199 - acc: 0.0363 - ETA: 49s - loss: 4.6197 - acc: 0.0362 - ETA: 48s - loss: 4.6194 - acc: 0.0361 - ETA: 47s - loss: 4.6194 - acc: 0.0361 - ETA: 46s - loss: 4.6183 - acc: 0.0360 - ETA: 45s - loss: 4.6179 - acc: 0.0359 - ETA: 44s - loss: 4.6175 - acc: 0.0363 - ETA: 43s - loss: 4.6172 - acc: 0.0363 - ETA: 43s - loss: 4.6191 - acc: 0.0362 - ETA: 42s - loss: 4.6191 - acc: 0.0361 - ETA: 41s - loss: 4.6201 - acc: 0.0359 - ETA: 40s - loss: 4.6202 - acc: 0.0358 - ETA: 39s - loss: 4.6194 - acc: 0.0357 - ETA: 38s - loss: 4.6201 - acc: 0.0356 - ETA: 37s - loss: 4.6200 - acc: 0.0354 - ETA: 36s - loss: 4.6209 - acc: 0.0355 - ETA: 35s - loss: 4.6207 - acc: 0.0355 - ETA: 34s - loss: 4.6211 - acc: 0.0356 - ETA: 34s - loss: 4.6210 - acc: 0.0358 - ETA: 33s - loss: 4.6214 - acc: 0.0360 - ETA: 32s - loss: 4.6216 - acc: 0.0359 - ETA: 31s - loss: 4.6219 - acc: 0.0358 - ETA: 30s - loss: 4.6220 - acc: 0.0360 - ETA: 29s - loss: 4.6212 - acc: 0.0360 - ETA: 28s - loss: 4.6214 - acc: 0.0359 - ETA: 27s - loss: 4.6216 - acc: 0.0360 - ETA: 26s - loss: 4.6226 - acc: 0.0359 - ETA: 26s - loss: 4.6228 - acc: 0.0359 - ETA: 25s - loss: 4.6227 - acc: 0.0358 - ETA: 24s - loss: 4.6225 - acc: 0.0358 - ETA: 23s - loss: 4.6230 - acc: 0.0357 - ETA: 22s - loss: 4.6224 - acc: 0.0359 - ETA: 21s - loss: 4.6219 - acc: 0.0361 - ETA: 20s - loss: 4.6227 - acc: 0.0360 - ETA: 19s - loss: 4.6235 - acc: 0.0359 - ETA: 18s - loss: 4.6236 - acc: 0.0358 - ETA: 17s - loss: 4.6234 - acc: 0.0358 - ETA: 17s - loss: 4.6230 - acc: 0.0357 - ETA: 16s - loss: 4.6229 - acc: 0.0356 - ETA: 15s - loss: 4.6221 - acc: 0.0356 - ETA: 14s - loss: 4.6217 - acc: 0.0357 - ETA: 13s - loss: 4.6214 - acc: 0.0356 - ETA: 12s - loss: 4.6209 - acc: 0.0356 - ETA: 11s - loss: 4.6211 - acc: 0.0357 - ETA: 10s - loss: 4.6218 - acc: 0.0357 - ETA: 9s - loss: 4.6222 - acc: 0.0358  - ETA: 8s - loss: 4.6219 - acc: 0.0356 - ETA: 8s - loss: 4.6220 - acc: 0.0357 - ETA: 7s - loss: 4.6211 - acc: 0.0357 - ETA: 6s - loss: 4.6215 - acc: 0.0356 - ETA: 5s - loss: 4.6223 - acc: 0.0355 - ETA: 4s - loss: 4.6228 - acc: 0.0356 - ETA: 3s - loss: 4.6223 - acc: 0.0356 - ETA: 2s - loss: 4.6223 - acc: 0.0356 - ETA: 1s - loss: 4.6225 - acc: 0.0355 - ETA: 0s - loss: 4.6228 - acc: 0.0354Epoch 00009: val_loss improved from 4.69424 to 4.67962, saving model to weights.best.from_scratch.hdf5
    6680/6680 [==============================] - 315s - loss: 4.6223 - acc: 0.0355 - val_loss: 4.6796 - val_acc: 0.0347
    --- Sction Complete ----
    

    Load the Model with the Best Validation Loss

    In [16]:
    model.load_weights('weights.best.from_scratch.hdf5')
    

    Test the Model

    Try out your model on the test dataset of dog images. Ensure that your test accuracy is greater than 1%.

    In [17]:
    # get index of predicted dog breed for each image in test set
    dog_breed_predictions = [np.argmax(model.predict(np.expand_dims(tensor, axis=0))) for tensor in test_tensors]
    
    # report test accuracy
    test_accuracy = 100*np.sum(np.array(dog_breed_predictions)==np.argmax(test_targets, axis=1))/len(dog_breed_predictions)
    print('Test accuracy: %.4f%%' % test_accuracy)
    
    Test accuracy: 3.3493%
    

    Step 4: Use a CNN to Classify Dog Breeds

    To reduce training time without sacrificing accuracy, we show you how to train a CNN using transfer learning. In the following step, you will get a chance to use transfer learning to train your own CNN.

    Obtain Bottleneck Features

    In [18]:
    bottleneck_features = np.load('C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/DogVGG16Data.npz')
    #bottleneck_features = np.load('bottleneck_features/DogVGG16Data.npz')
    train_VGG16 = bottleneck_features['train']
    valid_VGG16 = bottleneck_features['valid']
    test_VGG16 = bottleneck_features['test']
    

    Model Architecture

    The model uses the the pre-trained VGG-16 model as a fixed feature extractor, where the last convolutional output of VGG-16 is fed as input to our model. We only add a global average pooling layer and a fully connected layer, where the latter contains one node for each dog category and is equipped with a softmax.

    In [19]:
    VGG16_model = Sequential()
    VGG16_model.add(GlobalAveragePooling2D(input_shape=train_VGG16.shape[1:]))
    VGG16_model.add(Dense(133, activation='softmax'))
    
    VGG16_model.summary()
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    global_average_pooling2d_2 ( (None, 512)               0         
    _________________________________________________________________
    dense_2 (Dense)              (None, 133)               68229     
    =================================================================
    Total params: 68,229.0
    Trainable params: 68,229.0
    Non-trainable params: 0.0
    _________________________________________________________________
    

    Compile the Model

    In [20]:
    VGG16_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
    

    Train the Model

    In [21]:
    checkpointer = ModelCheckpoint(filepath='weights.best.VGG16.hdf5', 
                                   verbose=1, save_best_only=True)
    
    VGG16_model.fit(train_VGG16, train_targets, 
              validation_data=(valid_VGG16, valid_targets),
              epochs=20, batch_size=20, callbacks=[checkpointer], verbose=1)
    
    print('---I am done saving model VGG16----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/20
    6640/6680 [============================>.] - ETA: 225s - loss: 16.1181 - acc: 0.0000e+00 - ETA: 48s - loss: 15.0262 - acc: 0.0100      - ETA: 21s - loss: 14.9445 - acc: 0.0292 - ETA: 14s - loss: 14.6817 - acc: 0.0306 - ETA: 11s - loss: 14.7135 - acc: 0.0220 - ETA: 8s - loss: 14.6312 - acc: 0.0250  - ETA: 7s - loss: 14.5570 - acc: 0.0275 - ETA: 6s - loss: 14.3615 - acc: 0.0330 - ETA: 5s - loss: 14.3363 - acc: 0.0345 - ETA: 5s - loss: 14.1580 - acc: 0.0371 - ETA: 4s - loss: 14.1395 - acc: 0.0391 - ETA: 4s - loss: 14.1271 - acc: 0.0395 - ETA: 3s - loss: 14.0813 - acc: 0.0422 - ETA: 3s - loss: 13.9699 - acc: 0.0456 - ETA: 3s - loss: 13.9455 - acc: 0.0480 - ETA: 3s - loss: 13.8679 - acc: 0.0505 - ETA: 2s - loss: 13.7236 - acc: 0.0561 - ETA: 2s - loss: 13.6687 - acc: 0.0574 - ETA: 2s - loss: 13.5712 - acc: 0.0620 - ETA: 2s - loss: 13.5113 - acc: 0.0633 - ETA: 2s - loss: 13.4529 - acc: 0.0641 - ETA: 2s - loss: 13.3811 - acc: 0.0681 - ETA: 2s - loss: 13.3432 - acc: 0.0702 - ETA: 1s - loss: 13.3195 - acc: 0.0715 - ETA: 1s - loss: 13.2284 - acc: 0.0759 - ETA: 1s - loss: 13.1327 - acc: 0.0794 - ETA: 1s - loss: 13.0354 - acc: 0.0829 - ETA: 1s - loss: 12.9612 - acc: 0.0856 - ETA: 1s - loss: 12.9408 - acc: 0.0856 - ETA: 1s - loss: 12.8897 - acc: 0.0878 - ETA: 1s - loss: 12.8116 - acc: 0.0915 - ETA: 1s - loss: 12.7720 - acc: 0.0934 - ETA: 1s - loss: 12.7288 - acc: 0.0962 - ETA: 1s - loss: 12.6966 - acc: 0.0985 - ETA: 0s - loss: 12.6478 - acc: 0.1017 - ETA: 0s - loss: 12.6059 - acc: 0.1040 - ETA: 0s - loss: 12.5766 - acc: 0.1051 - ETA: 0s - loss: 12.5274 - acc: 0.1079 - ETA: 0s - loss: 12.4939 - acc: 0.1101 - ETA: 0s - loss: 12.4546 - acc: 0.1118 - ETA: 0s - loss: 12.4258 - acc: 0.1147 - ETA: 0s - loss: 12.3965 - acc: 0.1170 - ETA: 0s - loss: 12.3760 - acc: 0.1178 - ETA: 0s - loss: 12.3278 - acc: 0.1215 - ETA: 0s - loss: 12.2942 - acc: 0.1230 - ETA: 0s - loss: 12.2537 - acc: 0.1252 - ETA: 0s - loss: 12.2081 - acc: 0.1275 - ETA: 0s - loss: 12.1748 - acc: 0.1298Epoch 00000: val_loss improved from inf to 10.54641, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 3s - loss: 12.1716 - acc: 0.1301 - val_loss: 10.5464 - val_acc: 0.2347
    Epoch 2/20
    6620/6680 [============================>.] - ETA: 2s - loss: 9.7809 - acc: 0.3500 - ETA: 2s - loss: 10.1517 - acc: 0.2778 - ETA: 2s - loss: 10.6195 - acc: 0.2441 - ETA: 2s - loss: 10.3248 - acc: 0.2580 - ETA: 2s - loss: 10.4067 - acc: 0.2531 - ETA: 2s - loss: 10.3106 - acc: 0.2600 - ETA: 1s - loss: 10.1721 - acc: 0.2649 - ETA: 1s - loss: 10.1485 - acc: 0.2655 - ETA: 1s - loss: 10.1540 - acc: 0.2635 - ETA: 1s - loss: 10.1622 - acc: 0.2621 - ETA: 1s - loss: 10.0949 - acc: 0.2688 - ETA: 1s - loss: 10.1622 - acc: 0.2659 - ETA: 1s - loss: 10.1103 - acc: 0.2672 - ETA: 1s - loss: 10.1752 - acc: 0.2645 - ETA: 1s - loss: 10.1431 - acc: 0.2671 - ETA: 1s - loss: 10.1451 - acc: 0.2661 - ETA: 1s - loss: 10.1443 - acc: 0.2650 - ETA: 1s - loss: 10.1202 - acc: 0.2673 - ETA: 1s - loss: 10.1176 - acc: 0.2690 - ETA: 1s - loss: 10.1507 - acc: 0.2686 - ETA: 1s - loss: 10.1555 - acc: 0.2697 - ETA: 1s - loss: 10.1494 - acc: 0.2711 - ETA: 1s - loss: 10.1079 - acc: 0.2747 - ETA: 1s - loss: 10.1084 - acc: 0.2749 - ETA: 1s - loss: 10.0884 - acc: 0.2773 - ETA: 1s - loss: 10.1189 - acc: 0.2767 - ETA: 0s - loss: 10.1247 - acc: 0.2783 - ETA: 0s - loss: 10.1228 - acc: 0.2783 - ETA: 0s - loss: 10.0956 - acc: 0.2805 - ETA: 0s - loss: 10.1012 - acc: 0.2813 - ETA: 0s - loss: 10.0860 - acc: 0.2820 - ETA: 0s - loss: 10.0923 - acc: 0.2817 - ETA: 0s - loss: 10.0852 - acc: 0.2824 - ETA: 0s - loss: 10.0686 - acc: 0.2830 - ETA: 0s - loss: 10.0529 - acc: 0.2833 - ETA: 0s - loss: 10.0288 - acc: 0.2852 - ETA: 0s - loss: 10.0115 - acc: 0.2864 - ETA: 0s - loss: 10.0229 - acc: 0.2860 - ETA: 0s - loss: 10.0165 - acc: 0.2874 - ETA: 0s - loss: 10.0095 - acc: 0.2881 - ETA: 0s - loss: 9.9990 - acc: 0.2892  - ETA: 0s - loss: 9.9969 - acc: 0.2891 - ETA: 0s - loss: 10.0016 - acc: 0.2890 - ETA: 0s - loss: 9.9971 - acc: 0.2894  - ETA: 0s - loss: 9.9874 - acc: 0.2898 - ETA: 0s - loss: 9.9926 - acc: 0.2899Epoch 00001: val_loss improved from 10.54641 to 9.91412, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 10.0018 - acc: 0.2898 - val_loss: 9.9141 - val_acc: 0.3030
    Epoch 3/20
    6600/6680 [============================>.] - ETA: 2s - loss: 8.7459 - acc: 0.4000 - ETA: 2s - loss: 8.9237 - acc: 0.3778 - ETA: 2s - loss: 9.3131 - acc: 0.3563 - ETA: 2s - loss: 9.2465 - acc: 0.3563 - ETA: 2s - loss: 9.1970 - acc: 0.3629 - ETA: 2s - loss: 9.2020 - acc: 0.3658 - ETA: 2s - loss: 9.2076 - acc: 0.3667 - ETA: 2s - loss: 9.1910 - acc: 0.3625 - ETA: 1s - loss: 9.1423 - acc: 0.3678 - ETA: 1s - loss: 9.2315 - acc: 0.3636 - ETA: 1s - loss: 9.1759 - acc: 0.3712 - ETA: 1s - loss: 9.2468 - acc: 0.3688 - ETA: 1s - loss: 9.1948 - acc: 0.3736 - ETA: 1s - loss: 9.2621 - acc: 0.3707 - ETA: 1s - loss: 9.2821 - acc: 0.3678 - ETA: 1s - loss: 9.3196 - acc: 0.3644 - ETA: 1s - loss: 9.3868 - acc: 0.3600 - ETA: 1s - loss: 9.4213 - acc: 0.3590 - ETA: 1s - loss: 9.3617 - acc: 0.3640 - ETA: 1s - loss: 9.4335 - acc: 0.3599 - ETA: 1s - loss: 9.4626 - acc: 0.3591 - ETA: 1s - loss: 9.4874 - acc: 0.3580 - ETA: 1s - loss: 9.5176 - acc: 0.3567 - ETA: 1s - loss: 9.4938 - acc: 0.3591 - ETA: 1s - loss: 9.5018 - acc: 0.3602 - ETA: 1s - loss: 9.4831 - acc: 0.3601 - ETA: 1s - loss: 9.4974 - acc: 0.3592 - ETA: 1s - loss: 9.5030 - acc: 0.3583 - ETA: 0s - loss: 9.4983 - acc: 0.3595 - ETA: 0s - loss: 9.5078 - acc: 0.3585 - ETA: 0s - loss: 9.4814 - acc: 0.3606 - ETA: 0s - loss: 9.4706 - acc: 0.3609 - ETA: 0s - loss: 9.4575 - acc: 0.3610 - ETA: 0s - loss: 9.4532 - acc: 0.3615 - ETA: 0s - loss: 9.4450 - acc: 0.3624 - ETA: 0s - loss: 9.4591 - acc: 0.3618 - ETA: 0s - loss: 9.4498 - acc: 0.3621 - ETA: 0s - loss: 9.4496 - acc: 0.3613 - ETA: 0s - loss: 9.4496 - acc: 0.3601 - ETA: 0s - loss: 9.4429 - acc: 0.3605 - ETA: 0s - loss: 9.4510 - acc: 0.3591 - ETA: 0s - loss: 9.4512 - acc: 0.3590 - ETA: 0s - loss: 9.4561 - acc: 0.3590 - ETA: 0s - loss: 9.4577 - acc: 0.3594 - ETA: 0s - loss: 9.4217 - acc: 0.3621 - ETA: 0s - loss: 9.4107 - acc: 0.3625 - ETA: 0s - loss: 9.4101 - acc: 0.3629Epoch 00002: val_loss improved from 9.91412 to 9.51400, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 9.4291 - acc: 0.3620 - val_loss: 9.5140 - val_acc: 0.3413
    Epoch 4/20
    6620/6680 [============================>.] - ETA: 2s - loss: 9.1227 - acc: 0.4000 - ETA: 2s - loss: 9.9239 - acc: 0.3625 - ETA: 2s - loss: 9.6257 - acc: 0.3700 - ETA: 2s - loss: 9.8077 - acc: 0.3591 - ETA: 2s - loss: 9.4547 - acc: 0.3741 - ETA: 2s - loss: 9.0714 - acc: 0.3958 - ETA: 2s - loss: 9.2616 - acc: 0.3872 - ETA: 2s - loss: 9.1280 - acc: 0.3940 - ETA: 2s - loss: 9.0648 - acc: 0.4000 - ETA: 1s - loss: 9.0857 - acc: 0.3984 - ETA: 1s - loss: 9.1321 - acc: 0.3958 - ETA: 1s - loss: 9.1310 - acc: 0.3962 - ETA: 1s - loss: 9.1157 - acc: 0.3977 - ETA: 1s - loss: 9.0406 - acc: 0.4005 - ETA: 1s - loss: 9.0146 - acc: 0.4010 - ETA: 1s - loss: 9.0636 - acc: 0.3972 - ETA: 1s - loss: 9.0225 - acc: 0.3996 - ETA: 1s - loss: 9.0700 - acc: 0.3964 - ETA: 1s - loss: 9.0579 - acc: 0.3969 - ETA: 1s - loss: 9.0108 - acc: 0.3989 - ETA: 1s - loss: 8.9717 - acc: 0.4017 - ETA: 1s - loss: 8.9703 - acc: 0.4020 - ETA: 1s - loss: 8.9462 - acc: 0.4038 - ETA: 1s - loss: 8.9346 - acc: 0.4042 - ETA: 1s - loss: 8.9498 - acc: 0.4031 - ETA: 1s - loss: 8.9779 - acc: 0.4008 - ETA: 1s - loss: 8.9970 - acc: 0.3995 - ETA: 0s - loss: 8.9898 - acc: 0.4005 - ETA: 0s - loss: 9.0074 - acc: 0.3998 - ETA: 0s - loss: 9.0407 - acc: 0.3974 - ETA: 0s - loss: 9.0566 - acc: 0.3965 - ETA: 0s - loss: 9.0614 - acc: 0.3964 - ETA: 0s - loss: 9.0496 - acc: 0.3968 - ETA: 0s - loss: 9.0469 - acc: 0.3971 - ETA: 0s - loss: 9.0503 - acc: 0.3973 - ETA: 0s - loss: 9.0490 - acc: 0.3978 - ETA: 0s - loss: 9.0533 - acc: 0.3973 - ETA: 0s - loss: 9.0648 - acc: 0.3962 - ETA: 0s - loss: 9.0477 - acc: 0.3971 - ETA: 0s - loss: 9.0367 - acc: 0.3982 - ETA: 0s - loss: 9.0169 - acc: 0.3991 - ETA: 0s - loss: 9.0084 - acc: 0.3995 - ETA: 0s - loss: 9.0164 - acc: 0.3990 - ETA: 0s - loss: 9.0247 - acc: 0.3984 - ETA: 0s - loss: 9.0267 - acc: 0.3983 - ETA: 0s - loss: 9.0413 - acc: 0.3969 - ETA: 0s - loss: 9.0257 - acc: 0.3977Epoch 00003: val_loss improved from 9.51400 to 9.34174, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 9.0333 - acc: 0.3975 - val_loss: 9.3417 - val_acc: 0.3485
    Epoch 5/20
    6540/6680 [============================>.] - ETA: 2s - loss: 10.4795 - acc: 0.3500 - ETA: 2s - loss: 9.6925 - acc: 0.3625  - ETA: 2s - loss: 9.6189 - acc: 0.3733 - ETA: 2s - loss: 9.5894 - acc: 0.3727 - ETA: 2s - loss: 9.4211 - acc: 0.3879 - ETA: 2s - loss: 9.4404 - acc: 0.3861 - ETA: 2s - loss: 9.3930 - acc: 0.3849 - ETA: 2s - loss: 9.4322 - acc: 0.3830 - ETA: 2s - loss: 9.4299 - acc: 0.3825 - ETA: 1s - loss: 9.3453 - acc: 0.3862 - ETA: 1s - loss: 9.2830 - acc: 0.3889 - ETA: 1s - loss: 9.2209 - acc: 0.3930 - ETA: 1s - loss: 9.2720 - acc: 0.3913 - ETA: 1s - loss: 9.3131 - acc: 0.3887 - ETA: 1s - loss: 9.2573 - acc: 0.3910 - ETA: 1s - loss: 9.2667 - acc: 0.3897 - ETA: 1s - loss: 9.1791 - acc: 0.3948 - ETA: 1s - loss: 9.1277 - acc: 0.3988 - ETA: 1s - loss: 9.1248 - acc: 0.4000 - ETA: 1s - loss: 9.0966 - acc: 0.4022 - ETA: 1s - loss: 9.0828 - acc: 0.4038 - ETA: 1s - loss: 9.0740 - acc: 0.4043 - ETA: 1s - loss: 9.0640 - acc: 0.4045 - ETA: 1s - loss: 9.0613 - acc: 0.4055 - ETA: 1s - loss: 8.9994 - acc: 0.4096 - ETA: 1s - loss: 8.9648 - acc: 0.4118 - ETA: 1s - loss: 8.9740 - acc: 0.4110 - ETA: 1s - loss: 8.9920 - acc: 0.4096 - ETA: 0s - loss: 8.9947 - acc: 0.4098 - ETA: 0s - loss: 8.9569 - acc: 0.4118 - ETA: 0s - loss: 8.9485 - acc: 0.4129 - ETA: 0s - loss: 9.0008 - acc: 0.4100 - ETA: 0s - loss: 9.0113 - acc: 0.4088 - ETA: 0s - loss: 8.9714 - acc: 0.4117 - ETA: 0s - loss: 8.9512 - acc: 0.4132 - ETA: 0s - loss: 8.9230 - acc: 0.4146 - ETA: 0s - loss: 8.9110 - acc: 0.4154 - ETA: 0s - loss: 8.9015 - acc: 0.4159 - ETA: 0s - loss: 8.9146 - acc: 0.4144 - ETA: 0s - loss: 8.9041 - acc: 0.4147 - ETA: 0s - loss: 8.8894 - acc: 0.4149 - ETA: 0s - loss: 8.8951 - acc: 0.4151 - ETA: 0s - loss: 8.8548 - acc: 0.4176 - ETA: 0s - loss: 8.8492 - acc: 0.4181 - ETA: 0s - loss: 8.8492 - acc: 0.4177 - ETA: 0s - loss: 8.8402 - acc: 0.4181 - ETA: 0s - loss: 8.8476 - acc: 0.4176Epoch 00004: val_loss improved from 9.34174 to 9.23094, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 8.8540 - acc: 0.4175 - val_loss: 9.2309 - val_acc: 0.3629
    Epoch 6/20
    6660/6680 [============================>.] - ETA: 2s - loss: 9.1317 - acc: 0.4000 - ETA: 2s - loss: 9.3978 - acc: 0.3889 - ETA: 2s - loss: 8.9797 - acc: 0.4219 - ETA: 2s - loss: 8.3174 - acc: 0.4630 - ETA: 2s - loss: 8.2879 - acc: 0.4667 - ETA: 2s - loss: 8.4755 - acc: 0.4541 - ETA: 2s - loss: 8.4714 - acc: 0.4511 - ETA: 2s - loss: 8.3788 - acc: 0.4567 - ETA: 2s - loss: 8.3710 - acc: 0.4585 - ETA: 1s - loss: 8.6122 - acc: 0.4439 - ETA: 1s - loss: 8.7024 - acc: 0.4375 - ETA: 1s - loss: 8.7821 - acc: 0.4323 - ETA: 1s - loss: 8.8336 - acc: 0.4291 - ETA: 1s - loss: 8.7846 - acc: 0.4333 - ETA: 1s - loss: 8.7559 - acc: 0.4365 - ETA: 1s - loss: 8.7002 - acc: 0.4389 - ETA: 1s - loss: 8.7198 - acc: 0.4365 - ETA: 1s - loss: 8.7393 - acc: 0.4361 - ETA: 1s - loss: 8.7603 - acc: 0.4341 - ETA: 1s - loss: 8.7376 - acc: 0.4346 - ETA: 1s - loss: 8.7551 - acc: 0.4332 - ETA: 1s - loss: 8.7736 - acc: 0.4323 - ETA: 1s - loss: 8.7633 - acc: 0.4328 - ETA: 1s - loss: 8.7932 - acc: 0.4299 - ETA: 1s - loss: 8.8308 - acc: 0.4272 - ETA: 1s - loss: 8.8586 - acc: 0.4258 - ETA: 1s - loss: 8.8652 - acc: 0.4257 - ETA: 1s - loss: 8.8310 - acc: 0.4276 - ETA: 0s - loss: 8.8431 - acc: 0.4264 - ETA: 0s - loss: 8.8201 - acc: 0.4276 - ETA: 0s - loss: 8.8036 - acc: 0.4292 - ETA: 0s - loss: 8.7972 - acc: 0.4295 - ETA: 0s - loss: 8.7743 - acc: 0.4294 - ETA: 0s - loss: 8.7996 - acc: 0.4273 - ETA: 0s - loss: 8.8231 - acc: 0.4260 - ETA: 0s - loss: 8.8095 - acc: 0.4271 - ETA: 0s - loss: 8.7942 - acc: 0.4280 - ETA: 0s - loss: 8.7834 - acc: 0.4290 - ETA: 0s - loss: 8.7531 - acc: 0.4316 - ETA: 0s - loss: 8.7626 - acc: 0.4310 - ETA: 0s - loss: 8.7823 - acc: 0.4300 - ETA: 0s - loss: 8.7505 - acc: 0.4316 - ETA: 0s - loss: 8.7454 - acc: 0.4319 - ETA: 0s - loss: 8.7455 - acc: 0.4320 - ETA: 0s - loss: 8.7281 - acc: 0.4329 - ETA: 0s - loss: 8.7173 - acc: 0.4337 - ETA: 0s - loss: 8.7216 - acc: 0.4331 - ETA: 0s - loss: 8.7237 - acc: 0.4330Epoch 00005: val_loss improved from 9.23094 to 9.13905, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 8.7253 - acc: 0.4329 - val_loss: 9.1390 - val_acc: 0.3617
    Epoch 7/20
    6520/6680 [============================>.] - ETA: 2s - loss: 5.9485 - acc: 0.6000 - ETA: 2s - loss: 7.5389 - acc: 0.4944 - ETA: 2s - loss: 7.5252 - acc: 0.5000 - ETA: 2s - loss: 8.0882 - acc: 0.4696 - ETA: 2s - loss: 8.2535 - acc: 0.4617 - ETA: 2s - loss: 8.1455 - acc: 0.4689 - ETA: 2s - loss: 8.1898 - acc: 0.4636 - ETA: 2s - loss: 8.1081 - acc: 0.4706 - ETA: 2s - loss: 8.2376 - acc: 0.4647 - ETA: 1s - loss: 8.2751 - acc: 0.4623 - ETA: 1s - loss: 8.2879 - acc: 0.4625 - ETA: 1s - loss: 8.2955 - acc: 0.4633 - ETA: 1s - loss: 8.3092 - acc: 0.4622 - ETA: 1s - loss: 8.2708 - acc: 0.4651 - ETA: 1s - loss: 8.3169 - acc: 0.4630 - ETA: 1s - loss: 8.3077 - acc: 0.4636 - ETA: 1s - loss: 8.3901 - acc: 0.4592 - ETA: 1s - loss: 8.4074 - acc: 0.4583 - ETA: 1s - loss: 8.4898 - acc: 0.4531 - ETA: 1s - loss: 8.4890 - acc: 0.4526 - ETA: 1s - loss: 8.4678 - acc: 0.4535 - ETA: 1s - loss: 8.4988 - acc: 0.4513 - ETA: 1s - loss: 8.4737 - acc: 0.4529 - ETA: 1s - loss: 8.5017 - acc: 0.4512 - ETA: 1s - loss: 8.4779 - acc: 0.4520 - ETA: 1s - loss: 8.4624 - acc: 0.4528 - ETA: 1s - loss: 8.4487 - acc: 0.4530 - ETA: 1s - loss: 8.4841 - acc: 0.4505 - ETA: 0s - loss: 8.4809 - acc: 0.4498 - ETA: 0s - loss: 8.4814 - acc: 0.4502 - ETA: 0s - loss: 8.4697 - acc: 0.4505 - ETA: 0s - loss: 8.4581 - acc: 0.4500 - ETA: 0s - loss: 8.4586 - acc: 0.4502 - ETA: 0s - loss: 8.4758 - acc: 0.4491 - ETA: 0s - loss: 8.4687 - acc: 0.4492 - ETA: 0s - loss: 8.4423 - acc: 0.4504 - ETA: 0s - loss: 8.4291 - acc: 0.4510 - ETA: 0s - loss: 8.4306 - acc: 0.4506 - ETA: 0s - loss: 8.4489 - acc: 0.4480 - ETA: 0s - loss: 8.4546 - acc: 0.4471 - ETA: 0s - loss: 8.4826 - acc: 0.4451 - ETA: 0s - loss: 8.4678 - acc: 0.4466 - ETA: 0s - loss: 8.4395 - acc: 0.4479 - ETA: 0s - loss: 8.4532 - acc: 0.4466 - ETA: 0s - loss: 8.4698 - acc: 0.4458 - ETA: 0s - loss: 8.4928 - acc: 0.4442Epoch 00006: val_loss improved from 9.13905 to 8.91883, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 8.5197 - acc: 0.4422 - val_loss: 8.9188 - val_acc: 0.3593
    Epoch 8/20
    6580/6680 [============================>.] - ETA: 2s - loss: 5.6706 - acc: 0.6500 - ETA: 2s - loss: 7.6052 - acc: 0.5000 - ETA: 2s - loss: 8.3374 - acc: 0.4633 - ETA: 2s - loss: 8.4189 - acc: 0.4545 - ETA: 2s - loss: 8.3928 - acc: 0.4534 - ETA: 2s - loss: 8.5324 - acc: 0.4431 - ETA: 2s - loss: 8.4775 - acc: 0.4420 - ETA: 2s - loss: 8.5052 - acc: 0.4431 - ETA: 2s - loss: 8.5769 - acc: 0.4404 - ETA: 1s - loss: 8.3954 - acc: 0.4500 - ETA: 1s - loss: 8.2935 - acc: 0.4535 - ETA: 1s - loss: 8.1918 - acc: 0.4588 - ETA: 1s - loss: 8.2733 - acc: 0.4511 - ETA: 1s - loss: 8.2299 - acc: 0.4505 - ETA: 1s - loss: 8.1957 - acc: 0.4520 - ETA: 1s - loss: 8.1905 - acc: 0.4509 - ETA: 1s - loss: 8.2271 - acc: 0.4491 - ETA: 1s - loss: 8.2201 - acc: 0.4508 - ETA: 1s - loss: 8.2439 - acc: 0.4496 - ETA: 1s - loss: 8.1559 - acc: 0.4559 - ETA: 1s - loss: 8.1909 - acc: 0.4531 - ETA: 1s - loss: 8.1200 - acc: 0.4573 - ETA: 1s - loss: 8.1164 - acc: 0.4579 - ETA: 1s - loss: 8.0717 - acc: 0.4599 - ETA: 1s - loss: 8.0769 - acc: 0.4601 - ETA: 1s - loss: 8.0535 - acc: 0.4625 - ETA: 1s - loss: 8.0245 - acc: 0.4644 - ETA: 1s - loss: 8.0376 - acc: 0.4637 - ETA: 0s - loss: 8.0399 - acc: 0.4632 - ETA: 0s - loss: 8.0714 - acc: 0.4615 - ETA: 0s - loss: 8.0951 - acc: 0.4600 - ETA: 0s - loss: 8.1031 - acc: 0.4595 - ETA: 0s - loss: 8.1142 - acc: 0.4580 - ETA: 0s - loss: 8.0760 - acc: 0.4605 - ETA: 0s - loss: 8.0641 - acc: 0.4620 - ETA: 0s - loss: 8.0474 - acc: 0.4625 - ETA: 0s - loss: 8.0433 - acc: 0.4631 - ETA: 0s - loss: 8.0381 - acc: 0.4643 - ETA: 0s - loss: 8.0182 - acc: 0.4659 - ETA: 0s - loss: 8.0396 - acc: 0.4650 - ETA: 0s - loss: 8.0423 - acc: 0.4653 - ETA: 0s - loss: 8.0178 - acc: 0.4665 - ETA: 0s - loss: 8.0351 - acc: 0.4651 - ETA: 0s - loss: 8.0410 - acc: 0.4649 - ETA: 0s - loss: 8.0533 - acc: 0.4644 - ETA: 0s - loss: 8.0628 - acc: 0.4643 - ETA: 0s - loss: 8.0670 - acc: 0.4644Epoch 00007: val_loss improved from 8.91883 to 8.52306, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 8.0725 - acc: 0.4644 - val_loss: 8.5231 - val_acc: 0.3892
    Epoch 9/20
    6660/6680 [============================>.] - ETA: 2s - loss: 8.8947 - acc: 0.4500 - ETA: 2s - loss: 8.1002 - acc: 0.4875 - ETA: 2s - loss: 7.6296 - acc: 0.5067 - ETA: 2s - loss: 7.9216 - acc: 0.4886 - ETA: 2s - loss: 7.8033 - acc: 0.4933 - ETA: 2s - loss: 7.9177 - acc: 0.4851 - ETA: 2s - loss: 7.9145 - acc: 0.4852 - ETA: 2s - loss: 7.9843 - acc: 0.4808 - ETA: 2s - loss: 7.9421 - acc: 0.4842 - ETA: 1s - loss: 7.8589 - acc: 0.4903 - ETA: 1s - loss: 7.8521 - acc: 0.4926 - ETA: 1s - loss: 7.8933 - acc: 0.4889 - ETA: 1s - loss: 7.8438 - acc: 0.4920 - ETA: 1s - loss: 7.8491 - acc: 0.4916 - ETA: 1s - loss: 7.8521 - acc: 0.4917 - ETA: 1s - loss: 7.8556 - acc: 0.4913 - ETA: 1s - loss: 7.8904 - acc: 0.4897 - ETA: 1s - loss: 7.8760 - acc: 0.4907 - ETA: 1s - loss: 7.9117 - acc: 0.4892 - ETA: 1s - loss: 7.8685 - acc: 0.4909 - ETA: 1s - loss: 7.8469 - acc: 0.4914 - ETA: 1s - loss: 7.8668 - acc: 0.4905 - ETA: 1s - loss: 7.8922 - acc: 0.4888 - ETA: 1s - loss: 7.9178 - acc: 0.4874 - ETA: 1s - loss: 7.9657 - acc: 0.4834 - ETA: 1s - loss: 7.9325 - acc: 0.4852 - ETA: 1s - loss: 7.9168 - acc: 0.4858 - ETA: 1s - loss: 7.9199 - acc: 0.4858 - ETA: 0s - loss: 7.8874 - acc: 0.4874 - ETA: 0s - loss: 7.8978 - acc: 0.4865 - ETA: 0s - loss: 7.8996 - acc: 0.4865 - ETA: 0s - loss: 7.9285 - acc: 0.4847 - ETA: 0s - loss: 7.9252 - acc: 0.4850 - ETA: 0s - loss: 7.9253 - acc: 0.4850 - ETA: 0s - loss: 7.9451 - acc: 0.4842 - ETA: 0s - loss: 7.9424 - acc: 0.4846 - ETA: 0s - loss: 7.9552 - acc: 0.4833 - ETA: 0s - loss: 7.9486 - acc: 0.4836 - ETA: 0s - loss: 7.9531 - acc: 0.4829 - ETA: 0s - loss: 7.9577 - acc: 0.4823 - ETA: 0s - loss: 7.9554 - acc: 0.4826 - ETA: 0s - loss: 7.9450 - acc: 0.4837 - ETA: 0s - loss: 7.9240 - acc: 0.4846 - ETA: 0s - loss: 7.9082 - acc: 0.4854 - ETA: 0s - loss: 7.9075 - acc: 0.4853 - ETA: 0s - loss: 7.8935 - acc: 0.4859 - ETA: 0s - loss: 7.8823 - acc: 0.4865Epoch 00008: val_loss improved from 8.52306 to 8.44016, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.8839 - acc: 0.4861 - val_loss: 8.4402 - val_acc: 0.4012
    Epoch 10/20
    6640/6680 [============================>.] - ETA: 2s - loss: 4.8366 - acc: 0.7000 - ETA: 2s - loss: 9.3440 - acc: 0.4062 - ETA: 2s - loss: 8.8731 - acc: 0.4300 - ETA: 2s - loss: 8.3799 - acc: 0.4636 - ETA: 2s - loss: 8.2034 - acc: 0.4741 - ETA: 2s - loss: 8.1923 - acc: 0.4750 - ETA: 2s - loss: 8.0851 - acc: 0.4826 - ETA: 2s - loss: 8.0130 - acc: 0.4860 - ETA: 2s - loss: 7.8975 - acc: 0.4939 - ETA: 2s - loss: 7.8890 - acc: 0.4945 - ETA: 1s - loss: 7.7692 - acc: 0.5021 - ETA: 1s - loss: 7.7745 - acc: 0.5000 - ETA: 1s - loss: 7.7870 - acc: 0.4988 - ETA: 1s - loss: 7.7501 - acc: 0.5022 - ETA: 1s - loss: 7.7508 - acc: 0.5020 - ETA: 1s - loss: 7.7411 - acc: 0.5009 - ETA: 1s - loss: 7.7231 - acc: 0.5026 - ETA: 1s - loss: 7.7195 - acc: 0.5029 - ETA: 1s - loss: 7.7280 - acc: 0.5020 - ETA: 1s - loss: 7.7796 - acc: 0.4993 - ETA: 1s - loss: 7.8117 - acc: 0.4972 - ETA: 1s - loss: 7.8509 - acc: 0.4953 - ETA: 1s - loss: 7.8474 - acc: 0.4955 - ETA: 1s - loss: 7.8658 - acc: 0.4942 - ETA: 1s - loss: 7.8566 - acc: 0.4944 - ETA: 1s - loss: 7.8846 - acc: 0.4921 - ETA: 1s - loss: 7.8349 - acc: 0.4957 - ETA: 1s - loss: 7.8379 - acc: 0.4948 - ETA: 1s - loss: 7.7987 - acc: 0.4970 - ETA: 0s - loss: 7.7772 - acc: 0.4983 - ETA: 0s - loss: 7.7816 - acc: 0.4981 - ETA: 0s - loss: 7.8018 - acc: 0.4973 - ETA: 0s - loss: 7.7890 - acc: 0.4967 - ETA: 0s - loss: 7.8133 - acc: 0.4955 - ETA: 0s - loss: 7.8092 - acc: 0.4954 - ETA: 0s - loss: 7.7773 - acc: 0.4974 - ETA: 0s - loss: 7.7819 - acc: 0.4972 - ETA: 0s - loss: 7.7866 - acc: 0.4971 - ETA: 0s - loss: 7.7804 - acc: 0.4976 - ETA: 0s - loss: 7.7823 - acc: 0.4973 - ETA: 0s - loss: 7.7801 - acc: 0.4973 - ETA: 0s - loss: 7.7849 - acc: 0.4969 - ETA: 0s - loss: 7.7871 - acc: 0.4968 - ETA: 0s - loss: 7.8077 - acc: 0.4954 - ETA: 0s - loss: 7.7969 - acc: 0.4960 - ETA: 0s - loss: 7.7852 - acc: 0.4970 - ETA: 0s - loss: 7.7845 - acc: 0.4974 - ETA: 0s - loss: 7.7643 - acc: 0.4986Epoch 00009: val_loss improved from 8.44016 to 8.38619, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.7542 - acc: 0.4994 - val_loss: 8.3862 - val_acc: 0.4048
    Epoch 11/20
    6540/6680 [============================>.] - ETA: 2s - loss: 6.4714 - acc: 0.6000 - ETA: 2s - loss: 6.1066 - acc: 0.6125 - ETA: 2s - loss: 7.1041 - acc: 0.5500 - ETA: 2s - loss: 7.1528 - acc: 0.5500 - ETA: 2s - loss: 7.3590 - acc: 0.5345 - ETA: 2s - loss: 7.3652 - acc: 0.5333 - ETA: 2s - loss: 7.4004 - acc: 0.5279 - ETA: 2s - loss: 7.3975 - acc: 0.5260 - ETA: 2s - loss: 7.4288 - acc: 0.5246 - ETA: 2s - loss: 7.5515 - acc: 0.5164 - ETA: 1s - loss: 7.5720 - acc: 0.5155 - ETA: 1s - loss: 7.5235 - acc: 0.5199 - ETA: 1s - loss: 7.5241 - acc: 0.5200 - ETA: 1s - loss: 7.4982 - acc: 0.5217 - ETA: 1s - loss: 7.4836 - acc: 0.5225 - ETA: 1s - loss: 7.5526 - acc: 0.5176 - ETA: 1s - loss: 7.5814 - acc: 0.5139 - ETA: 1s - loss: 7.6796 - acc: 0.5074 - ETA: 1s - loss: 7.6992 - acc: 0.5062 - ETA: 1s - loss: 7.7180 - acc: 0.5055 - ETA: 1s - loss: 7.7095 - acc: 0.5059 - ETA: 1s - loss: 7.6481 - acc: 0.5090 - ETA: 1s - loss: 7.5668 - acc: 0.5140 - ETA: 1s - loss: 7.5268 - acc: 0.5168 - ETA: 1s - loss: 7.5497 - acc: 0.5158 - ETA: 1s - loss: 7.5923 - acc: 0.5132 - ETA: 1s - loss: 7.6051 - acc: 0.5122 - ETA: 1s - loss: 7.6239 - acc: 0.5109 - ETA: 0s - loss: 7.5976 - acc: 0.5126 - ETA: 0s - loss: 7.5830 - acc: 0.5133 - ETA: 0s - loss: 7.5731 - acc: 0.5140 - ETA: 0s - loss: 7.5419 - acc: 0.5154 - ETA: 0s - loss: 7.5568 - acc: 0.5140 - ETA: 0s - loss: 7.5611 - acc: 0.5132 - ETA: 0s - loss: 7.5687 - acc: 0.5122 - ETA: 0s - loss: 7.5939 - acc: 0.5100 - ETA: 0s - loss: 7.5919 - acc: 0.5102 - ETA: 0s - loss: 7.6252 - acc: 0.5082 - ETA: 0s - loss: 7.6345 - acc: 0.5078 - ETA: 0s - loss: 7.6395 - acc: 0.5076 - ETA: 0s - loss: 7.6582 - acc: 0.5063 - ETA: 0s - loss: 7.6386 - acc: 0.5076 - ETA: 0s - loss: 7.6450 - acc: 0.5062 - ETA: 0s - loss: 7.6411 - acc: 0.5062 - ETA: 0s - loss: 7.6282 - acc: 0.5069 - ETA: 0s - loss: 7.6084 - acc: 0.5083 - ETA: 0s - loss: 7.6162 - acc: 0.5076Epoch 00010: val_loss improved from 8.38619 to 8.12543, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.6290 - acc: 0.5069 - val_loss: 8.1254 - val_acc: 0.4204
    Epoch 12/20
    6540/6680 [============================>.] - ETA: 2s - loss: 8.0595 - acc: 0.5000 - ETA: 2s - loss: 7.3829 - acc: 0.5278 - ETA: 2s - loss: 7.4089 - acc: 0.5281 - ETA: 2s - loss: 7.4770 - acc: 0.5196 - ETA: 2s - loss: 7.4631 - acc: 0.5183 - ETA: 2s - loss: 7.4256 - acc: 0.5230 - ETA: 2s - loss: 7.5039 - acc: 0.5182 - ETA: 2s - loss: 7.4778 - acc: 0.5176 - ETA: 2s - loss: 7.3535 - acc: 0.5259 - ETA: 1s - loss: 7.4328 - acc: 0.5215 - ETA: 1s - loss: 7.4494 - acc: 0.5222 - ETA: 1s - loss: 7.4447 - acc: 0.5209 - ETA: 1s - loss: 7.3822 - acc: 0.5238 - ETA: 1s - loss: 7.3914 - acc: 0.5237 - ETA: 1s - loss: 7.4539 - acc: 0.5205 - ETA: 1s - loss: 7.4851 - acc: 0.5182 - ETA: 1s - loss: 7.5330 - acc: 0.5158 - ETA: 1s - loss: 7.5291 - acc: 0.5165 - ETA: 1s - loss: 7.4821 - acc: 0.5191 - ETA: 1s - loss: 7.4771 - acc: 0.5193 - ETA: 1s - loss: 7.4817 - acc: 0.5187 - ETA: 1s - loss: 7.4837 - acc: 0.5178 - ETA: 1s - loss: 7.4801 - acc: 0.5183 - ETA: 1s - loss: 7.5192 - acc: 0.5160 - ETA: 1s - loss: 7.4922 - acc: 0.5176 - ETA: 1s - loss: 7.5217 - acc: 0.5164 - ETA: 1s - loss: 7.5274 - acc: 0.5163 - ETA: 1s - loss: 7.5488 - acc: 0.5147 - ETA: 1s - loss: 7.5243 - acc: 0.5162 - ETA: 0s - loss: 7.4984 - acc: 0.5178 - ETA: 0s - loss: 7.4587 - acc: 0.5196 - ETA: 0s - loss: 7.4476 - acc: 0.5202 - ETA: 0s - loss: 7.4542 - acc: 0.5198 - ETA: 0s - loss: 7.4759 - acc: 0.5186 - ETA: 0s - loss: 7.4504 - acc: 0.5202 - ETA: 0s - loss: 7.4631 - acc: 0.5189 - ETA: 0s - loss: 7.4866 - acc: 0.5176 - ETA: 0s - loss: 7.4690 - acc: 0.5184 - ETA: 0s - loss: 7.4702 - acc: 0.5186 - ETA: 0s - loss: 7.4917 - acc: 0.5174 - ETA: 0s - loss: 7.4713 - acc: 0.5188 - ETA: 0s - loss: 7.4936 - acc: 0.5176 - ETA: 0s - loss: 7.4758 - acc: 0.5179 - ETA: 0s - loss: 7.4896 - acc: 0.5173 - ETA: 0s - loss: 7.4967 - acc: 0.5168 - ETA: 0s - loss: 7.4893 - acc: 0.5175 - ETA: 0s - loss: 7.4916 - acc: 0.5176Epoch 00011: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 7.4776 - acc: 0.5183 - val_loss: 8.1265 - val_acc: 0.4156
    Epoch 13/20
    6540/6680 [============================>.] - ETA: 2s - loss: 7.3014 - acc: 0.5000 - ETA: 2s - loss: 7.1076 - acc: 0.5375 - ETA: 2s - loss: 7.3450 - acc: 0.5300 - ETA: 2s - loss: 7.4362 - acc: 0.5250 - ETA: 2s - loss: 7.3710 - acc: 0.5293 - ETA: 2s - loss: 7.4653 - acc: 0.5250 - ETA: 2s - loss: 7.4129 - acc: 0.5302 - ETA: 2s - loss: 7.4879 - acc: 0.5270 - ETA: 2s - loss: 7.4352 - acc: 0.5298 - ETA: 2s - loss: 7.5710 - acc: 0.5211 - ETA: 1s - loss: 7.4744 - acc: 0.5257 - ETA: 1s - loss: 7.4557 - acc: 0.5278 - ETA: 1s - loss: 7.4790 - acc: 0.5267 - ETA: 1s - loss: 7.4142 - acc: 0.5290 - ETA: 1s - loss: 7.3952 - acc: 0.5310 - ETA: 1s - loss: 7.3384 - acc: 0.5346 - ETA: 1s - loss: 7.3210 - acc: 0.5355 - ETA: 1s - loss: 7.3129 - acc: 0.5360 - ETA: 1s - loss: 7.3262 - acc: 0.5352 - ETA: 1s - loss: 7.3675 - acc: 0.5319 - ETA: 1s - loss: 7.3732 - acc: 0.5320 - ETA: 1s - loss: 7.3894 - acc: 0.5315 - ETA: 1s - loss: 7.3920 - acc: 0.5315 - ETA: 1s - loss: 7.4032 - acc: 0.5306 - ETA: 1s - loss: 7.3650 - acc: 0.5332 - ETA: 1s - loss: 7.3765 - acc: 0.5328 - ETA: 1s - loss: 7.3652 - acc: 0.5329 - ETA: 1s - loss: 7.3491 - acc: 0.5332 - ETA: 0s - loss: 7.3564 - acc: 0.5331 - ETA: 0s - loss: 7.3628 - acc: 0.5322 - ETA: 0s - loss: 7.3870 - acc: 0.5305 - ETA: 0s - loss: 7.3709 - acc: 0.5311 - ETA: 0s - loss: 7.3571 - acc: 0.5323 - ETA: 0s - loss: 7.3788 - acc: 0.5309 - ETA: 0s - loss: 7.3895 - acc: 0.5302 - ETA: 0s - loss: 7.4006 - acc: 0.5296 - ETA: 0s - loss: 7.4038 - acc: 0.5290 - ETA: 0s - loss: 7.3858 - acc: 0.5303 - ETA: 0s - loss: 7.3624 - acc: 0.5319 - ETA: 0s - loss: 7.3859 - acc: 0.5306 - ETA: 0s - loss: 7.3850 - acc: 0.5302 - ETA: 0s - loss: 7.3848 - acc: 0.5300 - ETA: 0s - loss: 7.3854 - acc: 0.5299 - ETA: 0s - loss: 7.3788 - acc: 0.5302 - ETA: 0s - loss: 7.3892 - acc: 0.5297 - ETA: 0s - loss: 7.3983 - acc: 0.5288 - ETA: 0s - loss: 7.4132 - acc: 0.5277Epoch 00012: val_loss improved from 8.12543 to 8.08215, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.4336 - acc: 0.5263 - val_loss: 8.0821 - val_acc: 0.4240
    Epoch 14/20
    6620/6680 [============================>.] - ETA: 2s - loss: 8.0941 - acc: 0.5000 - ETA: 2s - loss: 7.6756 - acc: 0.5125 - ETA: 2s - loss: 7.4931 - acc: 0.5267 - ETA: 2s - loss: 7.6002 - acc: 0.5227 - ETA: 2s - loss: 7.6592 - acc: 0.5190 - ETA: 2s - loss: 7.6436 - acc: 0.5208 - ETA: 2s - loss: 7.5260 - acc: 0.5267 - ETA: 2s - loss: 7.4002 - acc: 0.5330 - ETA: 2s - loss: 7.4082 - acc: 0.5330 - ETA: 2s - loss: 7.2301 - acc: 0.5445 - ETA: 1s - loss: 7.3640 - acc: 0.5352 - ETA: 1s - loss: 7.3927 - acc: 0.5321 - ETA: 1s - loss: 7.3340 - acc: 0.5365 - ETA: 1s - loss: 7.2730 - acc: 0.5402 - ETA: 1s - loss: 7.3296 - acc: 0.5364 - ETA: 1s - loss: 7.3579 - acc: 0.5344 - ETA: 1s - loss: 7.4096 - acc: 0.5314 - ETA: 1s - loss: 7.4141 - acc: 0.5313 - ETA: 1s - loss: 7.4443 - acc: 0.5295 - ETA: 1s - loss: 7.4426 - acc: 0.5291 - ETA: 1s - loss: 7.4442 - acc: 0.5280 - ETA: 1s - loss: 7.4643 - acc: 0.5260 - ETA: 1s - loss: 7.4886 - acc: 0.5242 - ETA: 1s - loss: 7.4677 - acc: 0.5244 - ETA: 1s - loss: 7.4641 - acc: 0.5251 - ETA: 1s - loss: 7.4570 - acc: 0.5256 - ETA: 1s - loss: 7.4051 - acc: 0.5284 - ETA: 1s - loss: 7.3886 - acc: 0.5292 - ETA: 1s - loss: 7.3955 - acc: 0.5277 - ETA: 0s - loss: 7.4001 - acc: 0.5275 - ETA: 0s - loss: 7.4157 - acc: 0.5265 - ETA: 0s - loss: 7.4603 - acc: 0.5236 - ETA: 0s - loss: 7.4737 - acc: 0.5224 - ETA: 0s - loss: 7.4221 - acc: 0.5254 - ETA: 0s - loss: 7.4006 - acc: 0.5266 - ETA: 0s - loss: 7.3817 - acc: 0.5278 - ETA: 0s - loss: 7.3885 - acc: 0.5271 - ETA: 0s - loss: 7.3730 - acc: 0.5277 - ETA: 0s - loss: 7.3504 - acc: 0.5292 - ETA: 0s - loss: 7.3743 - acc: 0.5276 - ETA: 0s - loss: 7.3803 - acc: 0.5272 - ETA: 0s - loss: 7.3809 - acc: 0.5267 - ETA: 0s - loss: 7.3794 - acc: 0.5264 - ETA: 0s - loss: 7.3609 - acc: 0.5276 - ETA: 0s - loss: 7.3669 - acc: 0.5273 - ETA: 0s - loss: 7.3631 - acc: 0.5275 - ETA: 0s - loss: 7.3631 - acc: 0.5276 - ETA: 0s - loss: 7.3676 - acc: 0.5273Epoch 00013: val_loss improved from 8.08215 to 7.92109, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.3757 - acc: 0.5268 - val_loss: 7.9211 - val_acc: 0.4335
    Epoch 15/20
    6660/6680 [============================>.] - ETA: 2s - loss: 8.0597 - acc: 0.5000 - ETA: 2s - loss: 7.0553 - acc: 0.5625 - ETA: 2s - loss: 7.1061 - acc: 0.5500 - ETA: 2s - loss: 7.0452 - acc: 0.5543 - ETA: 2s - loss: 6.9602 - acc: 0.5600 - ETA: 2s - loss: 7.2776 - acc: 0.5419 - ETA: 2s - loss: 7.2523 - acc: 0.5443 - ETA: 2s - loss: 7.1849 - acc: 0.5461 - ETA: 1s - loss: 7.1890 - acc: 0.5441 - ETA: 1s - loss: 7.2405 - acc: 0.5379 - ETA: 1s - loss: 7.2221 - acc: 0.5390 - ETA: 1s - loss: 7.1949 - acc: 0.5413 - ETA: 1s - loss: 7.2003 - acc: 0.5420 - ETA: 1s - loss: 7.1085 - acc: 0.5458 - ETA: 1s - loss: 7.1930 - acc: 0.5412 - ETA: 1s - loss: 7.1565 - acc: 0.5431 - ETA: 1s - loss: 7.1193 - acc: 0.5453 - ETA: 1s - loss: 7.1751 - acc: 0.5423 - ETA: 1s - loss: 7.2107 - acc: 0.5408 - ETA: 1s - loss: 7.1719 - acc: 0.5438 - ETA: 1s - loss: 7.2445 - acc: 0.5396 - ETA: 1s - loss: 7.2175 - acc: 0.5411 - ETA: 1s - loss: 7.1996 - acc: 0.5424 - ETA: 1s - loss: 7.1920 - acc: 0.5427 - ETA: 1s - loss: 7.2182 - acc: 0.5401 - ETA: 1s - loss: 7.2020 - acc: 0.5402 - ETA: 1s - loss: 7.1927 - acc: 0.5395 - ETA: 1s - loss: 7.1933 - acc: 0.5391 - ETA: 0s - loss: 7.2114 - acc: 0.5383 - ETA: 0s - loss: 7.2105 - acc: 0.5384 - ETA: 0s - loss: 7.1625 - acc: 0.5411 - ETA: 0s - loss: 7.1851 - acc: 0.5398 - ETA: 0s - loss: 7.2071 - acc: 0.5386 - ETA: 0s - loss: 7.2043 - acc: 0.5383 - ETA: 0s - loss: 7.2029 - acc: 0.5384 - ETA: 0s - loss: 7.1919 - acc: 0.5392 - ETA: 0s - loss: 7.1842 - acc: 0.5391 - ETA: 0s - loss: 7.2045 - acc: 0.5376 - ETA: 0s - loss: 7.1925 - acc: 0.5387 - ETA: 0s - loss: 7.2069 - acc: 0.5381 - ETA: 0s - loss: 7.2069 - acc: 0.5377 - ETA: 0s - loss: 7.2085 - acc: 0.5376 - ETA: 0s - loss: 7.2135 - acc: 0.5372 - ETA: 0s - loss: 7.2098 - acc: 0.5375 - ETA: 0s - loss: 7.1981 - acc: 0.5381 - ETA: 0s - loss: 7.1909 - acc: 0.5386 - ETA: 0s - loss: 7.1996 - acc: 0.5382 - ETA: 0s - loss: 7.1673 - acc: 0.5404Epoch 00014: val_loss improved from 7.92109 to 7.86037, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.1618 - acc: 0.5407 - val_loss: 7.8604 - val_acc: 0.4491
    Epoch 16/20
    6620/6680 [============================>.] - ETA: 2s - loss: 8.9890 - acc: 0.4000 - ETA: 2s - loss: 7.2796 - acc: 0.5375 - ETA: 2s - loss: 7.2819 - acc: 0.5367 - ETA: 2s - loss: 7.3100 - acc: 0.5386 - ETA: 2s - loss: 7.1269 - acc: 0.5500 - ETA: 2s - loss: 7.0027 - acc: 0.5554 - ETA: 2s - loss: 7.0909 - acc: 0.5489 - ETA: 2s - loss: 7.1438 - acc: 0.5461 - ETA: 2s - loss: 7.1904 - acc: 0.5440 - ETA: 1s - loss: 7.2594 - acc: 0.5408 - ETA: 1s - loss: 7.2044 - acc: 0.5444 - ETA: 1s - loss: 7.1826 - acc: 0.5449 - ETA: 1s - loss: 7.2076 - acc: 0.5442 - ETA: 1s - loss: 7.2760 - acc: 0.5403 - ETA: 1s - loss: 7.1922 - acc: 0.5455 - ETA: 1s - loss: 7.1238 - acc: 0.5495 - ETA: 1s - loss: 7.0808 - acc: 0.5522 - ETA: 1s - loss: 7.0969 - acc: 0.5517 - ETA: 1s - loss: 7.0587 - acc: 0.5535 - ETA: 1s - loss: 7.0580 - acc: 0.5537 - ETA: 1s - loss: 7.0490 - acc: 0.5535 - ETA: 1s - loss: 6.9974 - acc: 0.5567 - ETA: 1s - loss: 7.0001 - acc: 0.5564 - ETA: 1s - loss: 7.0574 - acc: 0.5521 - ETA: 1s - loss: 7.0902 - acc: 0.5503 - ETA: 1s - loss: 7.0694 - acc: 0.5520 - ETA: 1s - loss: 7.0713 - acc: 0.5514 - ETA: 1s - loss: 7.0828 - acc: 0.5508 - ETA: 1s - loss: 7.0630 - acc: 0.5518 - ETA: 0s - loss: 7.0871 - acc: 0.5500 - ETA: 0s - loss: 7.1038 - acc: 0.5483 - ETA: 0s - loss: 7.1018 - acc: 0.5486 - ETA: 0s - loss: 7.1162 - acc: 0.5478 - ETA: 0s - loss: 7.1450 - acc: 0.5461 - ETA: 0s - loss: 7.1488 - acc: 0.5460 - ETA: 0s - loss: 7.1076 - acc: 0.5480 - ETA: 0s - loss: 7.0740 - acc: 0.5504 - ETA: 0s - loss: 7.1019 - acc: 0.5489 - ETA: 0s - loss: 7.1079 - acc: 0.5483 - ETA: 0s - loss: 7.1320 - acc: 0.5469 - ETA: 0s - loss: 7.1385 - acc: 0.5466 - ETA: 0s - loss: 7.1242 - acc: 0.5477 - ETA: 0s - loss: 7.1106 - acc: 0.5485 - ETA: 0s - loss: 7.1031 - acc: 0.5490 - ETA: 0s - loss: 7.0923 - acc: 0.5498 - ETA: 0s - loss: 7.0965 - acc: 0.5497 - ETA: 0s - loss: 7.1113 - acc: 0.5486 - ETA: 0s - loss: 7.0907 - acc: 0.5498Epoch 00015: val_loss improved from 7.86037 to 7.79918, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.0875 - acc: 0.5501 - val_loss: 7.7992 - val_acc: 0.4455
    Epoch 17/20
    6580/6680 [============================>.] - ETA: 2s - loss: 5.6416 - acc: 0.6500 - ETA: 2s - loss: 6.7193 - acc: 0.5833 - ETA: 2s - loss: 6.8264 - acc: 0.5625 - ETA: 2s - loss: 6.8208 - acc: 0.5652 - ETA: 2s - loss: 6.7622 - acc: 0.5717 - ETA: 2s - loss: 6.8088 - acc: 0.5689 - ETA: 2s - loss: 6.9427 - acc: 0.5602 - ETA: 2s - loss: 7.0682 - acc: 0.5529 - ETA: 2s - loss: 7.0695 - acc: 0.5517 - ETA: 1s - loss: 7.2387 - acc: 0.5415 - ETA: 1s - loss: 7.2751 - acc: 0.5396 - ETA: 1s - loss: 7.1830 - acc: 0.5456 - ETA: 1s - loss: 7.2052 - acc: 0.5435 - ETA: 1s - loss: 7.2452 - acc: 0.5413 - ETA: 1s - loss: 7.2063 - acc: 0.5439 - ETA: 1s - loss: 7.1830 - acc: 0.5448 - ETA: 1s - loss: 7.1467 - acc: 0.5469 - ETA: 1s - loss: 7.1736 - acc: 0.5458 - ETA: 1s - loss: 7.1849 - acc: 0.5457 - ETA: 1s - loss: 7.2083 - acc: 0.5440 - ETA: 1s - loss: 7.1934 - acc: 0.5454 - ETA: 1s - loss: 7.1726 - acc: 0.5463 - ETA: 1s - loss: 7.2394 - acc: 0.5423 - ETA: 1s - loss: 7.2118 - acc: 0.5441 - ETA: 1s - loss: 7.2136 - acc: 0.5444 - ETA: 1s - loss: 7.1924 - acc: 0.5460 - ETA: 1s - loss: 7.1522 - acc: 0.5481 - ETA: 1s - loss: 7.1564 - acc: 0.5479 - ETA: 1s - loss: 7.1137 - acc: 0.5503 - ETA: 0s - loss: 7.1304 - acc: 0.5495 - ETA: 0s - loss: 7.1159 - acc: 0.5505 - ETA: 0s - loss: 7.0815 - acc: 0.5525 - ETA: 0s - loss: 7.0925 - acc: 0.5513 - ETA: 0s - loss: 7.0824 - acc: 0.5515 - ETA: 0s - loss: 7.1372 - acc: 0.5479 - ETA: 0s - loss: 7.1196 - acc: 0.5488 - ETA: 0s - loss: 7.1208 - acc: 0.5486 - ETA: 0s - loss: 7.1175 - acc: 0.5488 - ETA: 0s - loss: 7.1453 - acc: 0.5474 - ETA: 0s - loss: 7.1208 - acc: 0.5487 - ETA: 0s - loss: 7.1159 - acc: 0.5491 - ETA: 0s - loss: 7.1042 - acc: 0.5495 - ETA: 0s - loss: 7.1119 - acc: 0.5488 - ETA: 0s - loss: 7.1102 - acc: 0.5492 - ETA: 0s - loss: 7.1185 - acc: 0.5487 - ETA: 0s - loss: 7.0934 - acc: 0.5505 - ETA: 0s - loss: 7.0771 - acc: 0.5516 - ETA: 0s - loss: 7.0639 - acc: 0.5526Epoch 00016: val_loss improved from 7.79918 to 7.73562, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 7.0624 - acc: 0.5525 - val_loss: 7.7356 - val_acc: 0.4479
    Epoch 18/20
    6620/6680 [============================>.] - ETA: 2s - loss: 8.8650 - acc: 0.4500 - ETA: 2s - loss: 6.8963 - acc: 0.5722 - ETA: 2s - loss: 6.9022 - acc: 0.5719 - ETA: 2s - loss: 6.7415 - acc: 0.5783 - ETA: 2s - loss: 6.7443 - acc: 0.5790 - ETA: 2s - loss: 6.6052 - acc: 0.5882 - ETA: 2s - loss: 6.6246 - acc: 0.5867 - ETA: 2s - loss: 6.6348 - acc: 0.5856 - ETA: 1s - loss: 6.6682 - acc: 0.5831 - ETA: 1s - loss: 6.7033 - acc: 0.5803 - ETA: 1s - loss: 6.7674 - acc: 0.5767 - ETA: 1s - loss: 6.7800 - acc: 0.5762 - ETA: 1s - loss: 6.7663 - acc: 0.5747 - ETA: 1s - loss: 6.7769 - acc: 0.5745 - ETA: 1s - loss: 6.8022 - acc: 0.5733 - ETA: 1s - loss: 6.8566 - acc: 0.5699 - ETA: 1s - loss: 6.9037 - acc: 0.5665 - ETA: 1s - loss: 6.8593 - acc: 0.5689 - ETA: 1s - loss: 6.8467 - acc: 0.5698 - ETA: 1s - loss: 6.8562 - acc: 0.5688 - ETA: 1s - loss: 6.8375 - acc: 0.5696 - ETA: 1s - loss: 6.8539 - acc: 0.5683 - ETA: 1s - loss: 6.8358 - acc: 0.5697 - ETA: 1s - loss: 6.8882 - acc: 0.5668 - ETA: 1s - loss: 6.9224 - acc: 0.5646 - ETA: 1s - loss: 6.9278 - acc: 0.5640 - ETA: 1s - loss: 6.9455 - acc: 0.5627 - ETA: 1s - loss: 6.9532 - acc: 0.5622 - ETA: 1s - loss: 6.9886 - acc: 0.5598 - ETA: 0s - loss: 6.9515 - acc: 0.5621 - ETA: 0s - loss: 6.9693 - acc: 0.5610 - ETA: 0s - loss: 6.9418 - acc: 0.5630 - ETA: 0s - loss: 6.9484 - acc: 0.5626 - ETA: 0s - loss: 6.9399 - acc: 0.5631 - ETA: 0s - loss: 6.9360 - acc: 0.5633 - ETA: 0s - loss: 6.9485 - acc: 0.5626 - ETA: 0s - loss: 6.9704 - acc: 0.5612 - ETA: 0s - loss: 7.0312 - acc: 0.5573 - ETA: 0s - loss: 7.0556 - acc: 0.5556 - ETA: 0s - loss: 7.0464 - acc: 0.5560 - ETA: 0s - loss: 7.0202 - acc: 0.5578 - ETA: 0s - loss: 7.0299 - acc: 0.5569 - ETA: 0s - loss: 7.0299 - acc: 0.5571 - ETA: 0s - loss: 7.0300 - acc: 0.5569 - ETA: 0s - loss: 7.0455 - acc: 0.5561 - ETA: 0s - loss: 7.0408 - acc: 0.5563 - ETA: 0s - loss: 7.0350 - acc: 0.5566 - ETA: 0s - loss: 7.0405 - acc: 0.5562Epoch 00017: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 7.0430 - acc: 0.5560 - val_loss: 7.7391 - val_acc: 0.4563
    Epoch 19/20
    6540/6680 [============================>.] - ETA: 2s - loss: 7.2538 - acc: 0.5500 - ETA: 2s - loss: 6.9822 - acc: 0.5625 - ETA: 2s - loss: 6.9478 - acc: 0.5667 - ETA: 2s - loss: 6.8566 - acc: 0.5727 - ETA: 2s - loss: 7.0885 - acc: 0.5569 - ETA: 2s - loss: 7.1778 - acc: 0.5514 - ETA: 2s - loss: 7.1347 - acc: 0.5547 - ETA: 2s - loss: 7.0763 - acc: 0.5540 - ETA: 2s - loss: 6.8220 - acc: 0.5693 - ETA: 2s - loss: 6.9248 - acc: 0.5633 - ETA: 1s - loss: 7.0107 - acc: 0.5570 - ETA: 1s - loss: 7.2195 - acc: 0.5449 - ETA: 1s - loss: 7.2141 - acc: 0.5453 - ETA: 1s - loss: 7.2398 - acc: 0.5436 - ETA: 1s - loss: 7.1980 - acc: 0.5461 - ETA: 1s - loss: 7.1207 - acc: 0.5509 - ETA: 1s - loss: 7.1219 - acc: 0.5513 - ETA: 1s - loss: 7.1581 - acc: 0.5492 - ETA: 1s - loss: 7.1786 - acc: 0.5477 - ETA: 1s - loss: 7.1196 - acc: 0.5511 - ETA: 1s - loss: 7.1326 - acc: 0.5503 - ETA: 1s - loss: 7.1397 - acc: 0.5497 - ETA: 1s - loss: 7.1498 - acc: 0.5487 - ETA: 1s - loss: 7.1267 - acc: 0.5500 - ETA: 1s - loss: 7.0894 - acc: 0.5523 - ETA: 1s - loss: 7.0300 - acc: 0.5559 - ETA: 1s - loss: 7.0436 - acc: 0.5548 - ETA: 1s - loss: 7.0128 - acc: 0.5570 - ETA: 0s - loss: 6.9619 - acc: 0.5602 - ETA: 0s - loss: 6.9530 - acc: 0.5608 - ETA: 0s - loss: 6.9170 - acc: 0.5626 - ETA: 0s - loss: 6.9172 - acc: 0.5626 - ETA: 0s - loss: 6.9218 - acc: 0.5622 - ETA: 0s - loss: 6.9624 - acc: 0.5600 - ETA: 0s - loss: 6.9589 - acc: 0.5601 - ETA: 0s - loss: 6.9788 - acc: 0.5588 - ETA: 0s - loss: 6.9579 - acc: 0.5603 - ETA: 0s - loss: 6.9547 - acc: 0.5602 - ETA: 0s - loss: 6.9658 - acc: 0.5596 - ETA: 0s - loss: 6.9676 - acc: 0.5595 - ETA: 0s - loss: 6.9608 - acc: 0.5600 - ETA: 0s - loss: 6.9961 - acc: 0.5577 - ETA: 0s - loss: 6.9824 - acc: 0.5584 - ETA: 0s - loss: 6.9629 - acc: 0.5595 - ETA: 0s - loss: 6.9698 - acc: 0.5591 - ETA: 0s - loss: 6.9587 - acc: 0.5598 - ETA: 0s - loss: 6.9771 - acc: 0.5587Epoch 00018: val_loss improved from 7.73562 to 7.67761, saving model to weights.best.VGG16.hdf5
    6680/6680 [==============================] - 2s - loss: 6.9754 - acc: 0.5590 - val_loss: 7.6776 - val_acc: 0.4587
    Epoch 20/20
    6620/6680 [============================>.] - ETA: 2s - loss: 5.5032 - acc: 0.6000 - ETA: 2s - loss: 5.9272 - acc: 0.6250 - ETA: 2s - loss: 6.5486 - acc: 0.5900 - ETA: 2s - loss: 6.6269 - acc: 0.5864 - ETA: 2s - loss: 6.9674 - acc: 0.5655 - ETA: 2s - loss: 7.0039 - acc: 0.5625 - ETA: 2s - loss: 7.1024 - acc: 0.5558 - ETA: 2s - loss: 7.1625 - acc: 0.5510 - ETA: 2s - loss: 7.2173 - acc: 0.5474 - ETA: 2s - loss: 6.9910 - acc: 0.5617 - ETA: 1s - loss: 6.9273 - acc: 0.5655 - ETA: 1s - loss: 6.9786 - acc: 0.5622 - ETA: 1s - loss: 6.9970 - acc: 0.5606 - ETA: 1s - loss: 7.0205 - acc: 0.5582 - ETA: 1s - loss: 7.0621 - acc: 0.5556 - ETA: 1s - loss: 7.0718 - acc: 0.5542 - ETA: 1s - loss: 7.1273 - acc: 0.5509 - ETA: 1s - loss: 7.1185 - acc: 0.5517 - ETA: 1s - loss: 7.0317 - acc: 0.5547 - ETA: 1s - loss: 7.0076 - acc: 0.5563 - ETA: 1s - loss: 7.0121 - acc: 0.5564 - ETA: 1s - loss: 7.0326 - acc: 0.5544 - ETA: 1s - loss: 7.0085 - acc: 0.5548 - ETA: 1s - loss: 6.9807 - acc: 0.5568 - ETA: 1s - loss: 7.0123 - acc: 0.5547 - ETA: 1s - loss: 6.9837 - acc: 0.5560 - ETA: 1s - loss: 6.9764 - acc: 0.5568 - ETA: 1s - loss: 6.9670 - acc: 0.5576 - ETA: 1s - loss: 6.9262 - acc: 0.5596 - ETA: 0s - loss: 6.9261 - acc: 0.5598 - ETA: 0s - loss: 6.8917 - acc: 0.5618 - ETA: 0s - loss: 6.9108 - acc: 0.5610 - ETA: 0s - loss: 6.8626 - acc: 0.5642 - ETA: 0s - loss: 6.8574 - acc: 0.5646 - ETA: 0s - loss: 6.8556 - acc: 0.5650 - ETA: 0s - loss: 6.8581 - acc: 0.5650 - ETA: 0s - loss: 6.8378 - acc: 0.5661 - ETA: 0s - loss: 6.8864 - acc: 0.5632 - ETA: 0s - loss: 6.8694 - acc: 0.5644 - ETA: 0s - loss: 6.8792 - acc: 0.5640 - ETA: 0s - loss: 6.8743 - acc: 0.5645 - ETA: 0s - loss: 6.8813 - acc: 0.5642 - ETA: 0s - loss: 6.8849 - acc: 0.5640 - ETA: 0s - loss: 6.8622 - acc: 0.5655 - ETA: 0s - loss: 6.8586 - acc: 0.5658 - ETA: 0s - loss: 6.8605 - acc: 0.5658 - ETA: 0s - loss: 6.8771 - acc: 0.5648 - ETA: 0s - loss: 6.8594 - acc: 0.5660Epoch 00019: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 6.8536 - acc: 0.5663 - val_loss: 7.6891 - val_acc: 0.4479
    ---I am done saving model VGG16----
    

    Load the Model with the Best Validation Loss

    In [23]:
    VGG16_model.load_weights('weights.best.VGG16.hdf5')
    

    Test the Model

    Now, we can use the CNN to test how well it identifies breed within our test dataset of dog images. We print the test accuracy below.

    In [24]:
    # get index of predicted dog breed for each image in test set
    VGG16_predictions = [np.argmax(VGG16_model.predict(np.expand_dims(feature, axis=0))) for feature in test_VGG16]
    
    # report test accuracy
    test_accuracy = 100*np.sum(np.array(VGG16_predictions)==np.argmax(test_targets, axis=1))/len(VGG16_predictions)
    print('Test accuracy: %.4f%%' % test_accuracy)
    
    Test accuracy: 45.8134%
    

    Predict Dog Breed with the Model

    In [25]:
    from extract_bottleneck_features import *
    
    def VGG16_predict_breed(img_path):
        # extract bottleneck features
        bottleneck_feature = extract_VGG16(path_to_tensor(img_path))
        # obtain predicted vector
        predicted_vector = VGG16_model.predict(bottleneck_feature)
        # return dog breed that is predicted by the model
        return dog_names[np.argmax(predicted_vector)]
    

    Step 5: Create a CNN to Classify Dog Breeds (using Transfer Learning)

    You will now use transfer learning to create a CNN that can identify dog breed from images. Your CNN must attain at least 60% accuracy on the test set.

    In Step 4, we used transfer learning to create a CNN using VGG-16 bottleneck features. In this section, you must use the bottleneck features from a different pre-trained model. To make things easier for you, we have pre-computed the features for all of the networks that are currently available in Keras:

    The files are encoded as such:

    Dog{network}Data.npz
    
    

    where {network}, in the above filename, can be one of VGG19, Resnet50, InceptionV3, or Xception. Pick one of the above architectures, download the corresponding bottleneck features, and store the downloaded file in the bottleneck_features/ folder in the repository.

    (IMPLEMENTATION) Obtain Bottleneck Features

    In the code block below, extract the bottleneck features corresponding to the train, test, and validation sets by running the following:

    bottleneck_features = np.load('bottleneck_features/Dog{network}Data.npz')
    train_{network} = bottleneck_features['train']
    valid_{network} = bottleneck_features['valid']
    test_{network} = bottleneck_features['test']
    In [26]:
    ### TODO: Obtain bottleneck features from another pre-trained CNN.
    bottleneck_features_VGG19 = np.load('C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/DogVGG19Data.npz')
    train_VGG19 = bottleneck_features_VGG19['train']
    valid_VGG19 = bottleneck_features_VGG19['valid']
    test_VGG19 = bottleneck_features_VGG19['test']
    print('-- Obtained -- VGG19 Bottlenect ---')
    
    bottleneck_features_Resnet50 = np.load('C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/DogResnet50Data.npz')
    train_Resnet50 = bottleneck_features_Resnet50['train']
    valid_Resnet50 = bottleneck_features_Resnet50['valid']
    test_Resnet50 = bottleneck_features_Resnet50['test']
    print('-- Obtained -- Resnet50 Bottlenect ---')
    
    
    bottleneck_features_InceptionV3 = np.load('C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/DogInceptionV3Data.npz')
    train_InceptionV3 = bottleneck_features_InceptionV3['train']
    valid_InceptionV3 = bottleneck_features_InceptionV3['valid']
    test_InceptionV3 = bottleneck_features_InceptionV3['test']
    print('-- Obtained -- InceptionV3 Bottlenect ---')
    
    
    bottleneck_features_Xception = np.load('C:/Training/udacity/AI_NanoDegree/Term2/3.Convolutional Neural Networks Videos/datasets/DogXceptionData.npz')
    train_Xception = bottleneck_features_Xception['train']
    valid_Xception = bottleneck_features_Xception['valid']
    test_Xception = bottleneck_features_Xception['test']
    print('-- Obtained -- Xception Bottlenect ---')
    
    -- Obtained -- VGG19 Bottlenect ---
    -- Obtained -- Resnet50 Bottlenect ---
    -- Obtained -- InceptionV3 Bottlenect ---
    -- Obtained -- Xception Bottlenect ---
    

    (IMPLEMENTATION) Model Architecture

    Create a CNN to classify dog breed. At the end of your code cell block, summarize the layers of your model by executing the line:

        <your model's name>.summary()
    
    

    Question 5: Outline the steps you took to get to your final CNN architecture and your reasoning at each step. Describe why you think the architecture is suitable for the current problem.

    Answer:
    To decide upon the final CNN architecture I have considered pre-computed keras features VGG16, VGG19 , Resnet50, InceptionV3 and Xception.
    Steps that I have considered is
    1. Get the bottleneck feature
    2. Build the Model Architecture
    3. Compile Model
    4 . Train the model
    5. Load the Model with the Best Validation Loss
    6. Test the Model
    7. Look for a model with best accuracy


    In my analysis Xception , Inception and Resnet provides Test accuracy higher than 60%
    Xception Test accuracy: 85.0478%
    Inception Test accuracy: 77.1531%
    Resnet50 Test accuracy: 81.2201%


    Based on the accuracy results I have selected Xception as the best model for this project.

    In [27]:
    ### TODO: Define your architecture.
    VGG19_model = Sequential()
    VGG19_model.add(GlobalAveragePooling2D(input_shape=train_VGG19.shape[1:]))
    VGG19_model.add(Dense(133, activation='softmax'))
    VGG19_model.summary()
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    global_average_pooling2d_3 ( (None, 512)               0         
    _________________________________________________________________
    dense_3 (Dense)              (None, 133)               68229     
    =================================================================
    Total params: 68,229.0
    Trainable params: 68,229.0
    Non-trainable params: 0.0
    _________________________________________________________________
    
    In [32]:
    ### TODO: Define your architecture.
    Resnet50_model = Sequential()
    Resnet50_model.add(GlobalAveragePooling2D(input_shape=train_Resnet50.shape[1:]))
    Resnet50_model.add(Dense(133, activation='softmax'))
    Resnet50_model.summary()
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    global_average_pooling2d_8 ( (None, 2048)              0         
    _________________________________________________________________
    dense_8 (Dense)              (None, 133)               272517    
    =================================================================
    Total params: 272,517.0
    Trainable params: 272,517.0
    Non-trainable params: 0.0
    _________________________________________________________________
    
    In [31]:
    ### TODO: Define your architecture.
    InceptionV3_model = Sequential()
    InceptionV3_model.add(GlobalAveragePooling2D(input_shape=train_InceptionV3.shape[1:]))
    InceptionV3_model.add(Dense(133, activation='softmax'))
    InceptionV3_model.summary()
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    global_average_pooling2d_7 ( (None, 2048)              0         
    _________________________________________________________________
    dense_7 (Dense)              (None, 133)               272517    
    =================================================================
    Total params: 272,517.0
    Trainable params: 272,517.0
    Non-trainable params: 0.0
    _________________________________________________________________
    
    In [33]:
    ### TODO: Define your architecture.
    Xception_model = Sequential()
    Xception_model.add(GlobalAveragePooling2D(input_shape=train_Xception.shape[1:]))
    Xception_model.add(Dense(133, activation='softmax'))
    Xception_model.summary()
    
    _________________________________________________________________
    Layer (type)                 Output Shape              Param #   
    =================================================================
    global_average_pooling2d_9 ( (None, 2048)              0         
    _________________________________________________________________
    dense_9 (Dense)              (None, 133)               272517    
    =================================================================
    Total params: 272,517.0
    Trainable params: 272,517.0
    Non-trainable params: 0.0
    _________________________________________________________________
    

    (IMPLEMENTATION) Compile the Model

    In [34]:
    ### TODO: Compile the model.
    VGG19_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
    print('-- Model Compiled ---')
    
    -- Model Compiled ---
    
    In [35]:
    Resnet50_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
    print('-- Model Compiled ---')
    
    -- Model Compiled ---
    
    In [36]:
    InceptionV3_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
    print('-- Model Compiled ---')
    
    -- Model Compiled ---
    
    In [37]:
    Xception_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
    print('-- Model Compiled ---')
    
    -- Model Compiled ---
    

    (IMPLEMENTATION) Train the Model

    Train your model in the code cell below. Use model checkpointing to save the model that attains the best validation loss.

    You are welcome to augment the training data, but this is not a requirement.

    In [38]:
    ### TODO: Train the model.
    checkpointer_vgg19 = ModelCheckpoint(filepath='weights.best.VGG19.hdf5', 
                                   verbose=1, save_best_only=True)
    
    VGG19_model.fit(train_VGG19, train_targets, 
              validation_data=(valid_VGG19, valid_targets),
              epochs=20, batch_size=20, callbacks=[checkpointer_vgg19], verbose=1)
    
    print('---I am done saving model VGG19----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/20
    6640/6680 [============================>.] - ETA: 698s - loss: 15.1269 - acc: 0.0000e+00 - ETA: 392s - loss: 14.7043 - acc: 0.0000e+00 - ETA: 288s - loss: 14.7453 - acc: 0.0000e+00 - ETA: 236s - loss: 14.4638 - acc: 0.0125     - ETA: 212s - loss: 14.5182 - acc: 0.0100 - ETA: 190s - loss: 14.4975 - acc: 0.0083 - ETA: 174s - loss: 14.4742 - acc: 0.0143 - ETA: 161s - loss: 14.4100 - acc: 0.0188 - ETA: 150s - loss: 14.3931 - acc: 0.0222 - ETA: 140s - loss: 14.4761 - acc: 0.0200 - ETA: 133s - loss: 14.4191 - acc: 0.0182 - ETA: 127s - loss: 14.4757 - acc: 0.0167 - ETA: 122s - loss: 14.4309 - acc: 0.0192 - ETA: 117s - loss: 14.3169 - acc: 0.0214 - ETA: 114s - loss: 14.3138 - acc: 0.0233 - ETA: 112s - loss: 14.3176 - acc: 0.0250 - ETA: 108s - loss: 14.2732 - acc: 0.0235 - ETA: 105s - loss: 14.3085 - acc: 0.0222 - ETA: 103s - loss: 14.3015 - acc: 0.0237 - ETA: 101s - loss: 14.3218 - acc: 0.0225 - ETA: 106s - loss: 14.3013 - acc: 0.0238 - ETA: 109s - loss: 14.2505 - acc: 0.0227 - ETA: 107s - loss: 14.2036 - acc: 0.0239 - ETA: 104s - loss: 14.1658 - acc: 0.0229 - ETA: 103s - loss: 14.1015 - acc: 0.0240 - ETA: 101s - loss: 14.0968 - acc: 0.0250 - ETA: 99s - loss: 14.0720 - acc: 0.0241  - ETA: 97s - loss: 14.0687 - acc: 0.0232 - ETA: 96s - loss: 13.9252 - acc: 0.0259 - ETA: 94s - loss: 13.9079 - acc: 0.0250 - ETA: 93s - loss: 13.8352 - acc: 0.0242 - ETA: 91s - loss: 13.7845 - acc: 0.0250 - ETA: 90s - loss: 13.7429 - acc: 0.0242 - ETA: 89s - loss: 13.7076 - acc: 0.0265 - ETA: 88s - loss: 13.6744 - acc: 0.0257 - ETA: 87s - loss: 13.6211 - acc: 0.0264 - ETA: 85s - loss: 13.6071 - acc: 0.0270 - ETA: 84s - loss: 13.6165 - acc: 0.0263 - ETA: 84s - loss: 13.5830 - acc: 0.0269 - ETA: 83s - loss: 13.5986 - acc: 0.0275 - ETA: 82s - loss: 13.6063 - acc: 0.0268 - ETA: 82s - loss: 13.5504 - acc: 0.0298 - ETA: 81s - loss: 13.4874 - acc: 0.0326 - ETA: 80s - loss: 13.4165 - acc: 0.0341 - ETA: 79s - loss: 13.3966 - acc: 0.0344 - ETA: 79s - loss: 13.4509 - acc: 0.0337 - ETA: 78s - loss: 13.4286 - acc: 0.0351 - ETA: 77s - loss: 13.4183 - acc: 0.0365 - ETA: 77s - loss: 13.4004 - acc: 0.0367 - ETA: 76s - loss: 13.3857 - acc: 0.0370 - ETA: 75s - loss: 13.3610 - acc: 0.0373 - ETA: 75s - loss: 13.3253 - acc: 0.0394 - ETA: 74s - loss: 13.3133 - acc: 0.0396 - ETA: 73s - loss: 13.3427 - acc: 0.0389 - ETA: 73s - loss: 13.3187 - acc: 0.0391 - ETA: 72s - loss: 13.3298 - acc: 0.0393 - ETA: 72s - loss: 13.2891 - acc: 0.0412 - ETA: 71s - loss: 13.2900 - acc: 0.0414 - ETA: 71s - loss: 13.2549 - acc: 0.0424 - ETA: 70s - loss: 13.2534 - acc: 0.0425 - ETA: 70s - loss: 13.2351 - acc: 0.0443 - ETA: 69s - loss: 13.2400 - acc: 0.0444 - ETA: 69s - loss: 13.2484 - acc: 0.0437 - ETA: 69s - loss: 13.2109 - acc: 0.0445 - ETA: 68s - loss: 13.1792 - acc: 0.0462 - ETA: 68s - loss: 13.1870 - acc: 0.0455 - ETA: 67s - loss: 13.1926 - acc: 0.0455 - ETA: 67s - loss: 13.1922 - acc: 0.0449 - ETA: 66s - loss: 13.1885 - acc: 0.0442 - ETA: 66s - loss: 13.1861 - acc: 0.0450 - ETA: 66s - loss: 13.1821 - acc: 0.0451 - ETA: 65s - loss: 13.1536 - acc: 0.0451 - ETA: 65s - loss: 13.1070 - acc: 0.0466 - ETA: 65s - loss: 13.1032 - acc: 0.0466 - ETA: 64s - loss: 13.0903 - acc: 0.0467 - ETA: 65s - loss: 13.0640 - acc: 0.0474 - ETA: 67s - loss: 13.0550 - acc: 0.0474 - ETA: 69s - loss: 13.0205 - acc: 0.0487 - ETA: 72s - loss: 13.0136 - acc: 0.0487 - ETA: 74s - loss: 12.9918 - acc: 0.0513 - ETA: 77s - loss: 12.9942 - acc: 0.0512 - ETA: 82s - loss: 12.9597 - acc: 0.0518 - ETA: 88s - loss: 12.9107 - acc: 0.0518 - ETA: 92s - loss: 12.8944 - acc: 0.0524 - ETA: 93s - loss: 12.8829 - acc: 0.0535 - ETA: 93s - loss: 12.8562 - acc: 0.0547 - ETA: 93s - loss: 12.8439 - acc: 0.0546 - ETA: 92s - loss: 12.8308 - acc: 0.0551 - ETA: 91s - loss: 12.7970 - acc: 0.0573 - ETA: 90s - loss: 12.7878 - acc: 0.0578 - ETA: 89s - loss: 12.7755 - acc: 0.0582 - ETA: 88s - loss: 12.7695 - acc: 0.0587 - ETA: 88s - loss: 12.7575 - acc: 0.0591 - ETA: 87s - loss: 12.7346 - acc: 0.0606 - ETA: 86s - loss: 12.7116 - acc: 0.0626 - ETA: 85s - loss: 12.6884 - acc: 0.0641 - ETA: 84s - loss: 12.6714 - acc: 0.0649 - ETA: 84s - loss: 12.6523 - acc: 0.0653 - ETA: 83s - loss: 12.6341 - acc: 0.0657 - ETA: 82s - loss: 12.6186 - acc: 0.0660 - ETA: 81s - loss: 12.6154 - acc: 0.0668 - ETA: 81s - loss: 12.5674 - acc: 0.0691 - ETA: 80s - loss: 12.5703 - acc: 0.0689 - ETA: 79s - loss: 12.5490 - acc: 0.0697 - ETA: 79s - loss: 12.5321 - acc: 0.0700 - ETA: 78s - loss: 12.5208 - acc: 0.0708 - ETA: 77s - loss: 12.4989 - acc: 0.0715 - ETA: 77s - loss: 12.4982 - acc: 0.0713 - ETA: 76s - loss: 12.4929 - acc: 0.0716 - ETA: 75s - loss: 12.4970 - acc: 0.0718 - ETA: 75s - loss: 12.4896 - acc: 0.0721 - ETA: 74s - loss: 12.4861 - acc: 0.0728 - ETA: 73s - loss: 12.4596 - acc: 0.0730 - ETA: 73s - loss: 12.4288 - acc: 0.0741 - ETA: 72s - loss: 12.4162 - acc: 0.0752 - ETA: 72s - loss: 12.3985 - acc: 0.0759 - ETA: 71s - loss: 12.3750 - acc: 0.0761 - ETA: 70s - loss: 12.3413 - acc: 0.0771 - ETA: 70s - loss: 12.3214 - acc: 0.0786 - ETA: 69s - loss: 12.3304 - acc: 0.0788 - ETA: 69s - loss: 12.3118 - acc: 0.0789 - ETA: 68s - loss: 12.2956 - acc: 0.0791 - ETA: 67s - loss: 12.3102 - acc: 0.0785 - ETA: 67s - loss: 12.3110 - acc: 0.0786 - ETA: 66s - loss: 12.3017 - acc: 0.0784 - ETA: 66s - loss: 12.2880 - acc: 0.0790 - ETA: 65s - loss: 12.2842 - acc: 0.0799 - ETA: 65s - loss: 12.2752 - acc: 0.0801 - ETA: 64s - loss: 12.2448 - acc: 0.0818 - ETA: 64s - loss: 12.2291 - acc: 0.0823 - ETA: 63s - loss: 12.2362 - acc: 0.0821 - ETA: 62s - loss: 12.2208 - acc: 0.0830 - ETA: 62s - loss: 12.1956 - acc: 0.0838 - ETA: 61s - loss: 12.1751 - acc: 0.0843 - ETA: 61s - loss: 12.1665 - acc: 0.0856 - ETA: 60s - loss: 12.1354 - acc: 0.0860 - ETA: 60s - loss: 12.1123 - acc: 0.0869 - ETA: 59s - loss: 12.1189 - acc: 0.0866 - ETA: 59s - loss: 12.1077 - acc: 0.0878 - ETA: 58s - loss: 12.1050 - acc: 0.0875 - ETA: 58s - loss: 12.1035 - acc: 0.0876 - ETA: 58s - loss: 12.0832 - acc: 0.0891 - ETA: 57s - loss: 12.0700 - acc: 0.0899 - ETA: 57s - loss: 12.0630 - acc: 0.0903 - ETA: 56s - loss: 12.0584 - acc: 0.0907 - ETA: 56s - loss: 12.0586 - acc: 0.0904 - ETA: 55s - loss: 12.0418 - acc: 0.0915 - ETA: 55s - loss: 12.0163 - acc: 0.0926 - ETA: 55s - loss: 11.9941 - acc: 0.0933 - ETA: 54s - loss: 11.9882 - acc: 0.0933 - ETA: 54s - loss: 11.9663 - acc: 0.0944 - ETA: 53s - loss: 11.9561 - acc: 0.0947 - ETA: 53s - loss: 11.9529 - acc: 0.0951 - ETA: 52s - loss: 11.9387 - acc: 0.0961 - ETA: 52s - loss: 11.9127 - acc: 0.0974 - ETA: 51s - loss: 11.8905 - acc: 0.0987 - ETA: 51s - loss: 11.8859 - acc: 0.0990 - ETA: 50s - loss: 11.8830 - acc: 0.0994 - ETA: 50s - loss: 11.8726 - acc: 0.1009 - ETA: 49s - loss: 11.8697 - acc: 0.1013 - ETA: 49s - loss: 11.8530 - acc: 0.1022 - ETA: 48s - loss: 11.8484 - acc: 0.1028 - ETA: 48s - loss: 11.8194 - acc: 0.1037 - ETA: 48s - loss: 11.8066 - acc: 0.1037 - ETA: 47s - loss: 11.7941 - acc: 0.1045 - ETA: 47s - loss: 11.7919 - acc: 0.1048 - ETA: 46s - loss: 11.7910 - acc: 0.1054 - ETA: 46s - loss: 11.7745 - acc: 0.1065 - ETA: 46s - loss: 11.7719 - acc: 0.1065 - ETA: 45s - loss: 11.7738 - acc: 0.1071 - ETA: 45s - loss: 11.7577 - acc: 0.1079 - ETA: 44s - loss: 11.7515 - acc: 0.1087 - ETA: 44s - loss: 11.7498 - acc: 0.1090 - ETA: 44s - loss: 11.7329 - acc: 0.1098 - ETA: 43s - loss: 11.7275 - acc: 0.1106 - ETA: 43s - loss: 11.7171 - acc: 0.1108 - ETA: 42s - loss: 11.6972 - acc: 0.1127 - ETA: 42s - loss: 11.6872 - acc: 0.1138 - ETA: 41s - loss: 11.6707 - acc: 0.1137 - ETA: 41s - loss: 11.6701 - acc: 0.1139 - ETA: 41s - loss: 11.6715 - acc: 0.1144 - ETA: 40s - loss: 11.6752 - acc: 0.1137 - ETA: 40s - loss: 11.6505 - acc: 0.1150 - ETA: 40s - loss: 11.6464 - acc: 0.1152 - ETA: 39s - loss: 11.6391 - acc: 0.1151 - ETA: 39s - loss: 11.6410 - acc: 0.1159 - ETA: 38s - loss: 11.6311 - acc: 0.1163 - ETA: 38s - loss: 11.6051 - acc: 0.1170 - ETA: 38s - loss: 11.6018 - acc: 0.1175 - ETA: 37s - loss: 11.5953 - acc: 0.1179 - ETA: 37s - loss: 11.5831 - acc: 0.1188 - ETA: 36s - loss: 11.5677 - acc: 0.1201 - ETA: 36s - loss: 11.5583 - acc: 0.1207 - ETA: 36s - loss: 11.5547 - acc: 0.1204 - ETA: 35s - loss: 11.5557 - acc: 0.1205 - ETA: 35s - loss: 11.5326 - acc: 0.1222 - ETA: 35s - loss: 11.5178 - acc: 0.1228 - ETA: 34s - loss: 11.5057 - acc: 0.1235 - ETA: 34s - loss: 11.5079 - acc: 0.1234 - ETA: 34s - loss: 11.4973 - acc: 0.1240 - ETA: 33s - loss: 11.4824 - acc: 0.1249 - ETA: 33s - loss: 11.4756 - acc: 0.1248 - ETA: 33s - loss: 11.4569 - acc: 0.1251 - ETA: 32s - loss: 11.4430 - acc: 0.1265 - ETA: 32s - loss: 11.4335 - acc: 0.1268 - ETA: 32s - loss: 11.4378 - acc: 0.1272 - ETA: 31s - loss: 11.4440 - acc: 0.1268 - ETA: 31s - loss: 11.4360 - acc: 0.1276 - ETA: 31s - loss: 11.4264 - acc: 0.1287 - ETA: 30s - loss: 11.4182 - acc: 0.1288 - ETA: 30s - loss: 11.4196 - acc: 0.1291 - ETA: 30s - loss: 11.4114 - acc: 0.1295 - ETA: 29s - loss: 11.3874 - acc: 0.1308 - ETA: 29s - loss: 11.3698 - acc: 0.1315 - ETA: 29s - loss: 11.3606 - acc: 0.1323 - ETA: 28s - loss: 11.3650 - acc: 0.1326 - ETA: 28s - loss: 11.3698 - acc: 0.1329 - ETA: 28s - loss: 11.3667 - acc: 0.1333 - ETA: 27s - loss: 11.3557 - acc: 0.1333 - ETA: 27s - loss: 11.3547 - acc: 0.1334 - ETA: 27s - loss: 11.3439 - acc: 0.1346 - ETA: 26s - loss: 11.3513 - acc: 0.1345 - ETA: 26s - loss: 11.3523 - acc: 0.1343 - ETA: 26s - loss: 11.3404 - acc: 0.1355 - ETA: 25s - loss: 11.3309 - acc: 0.1362 - ETA: 25s - loss: 11.3144 - acc: 0.1369 - ETA: 24s - loss: 11.2793 - acc: 0.1388 - ETA: 24s - loss: 11.2665 - acc: 0.1393 - ETA: 24s - loss: 11.2581 - acc: 0.1398 - ETA: 24s - loss: 11.2618 - acc: 0.1398 - ETA: 23s - loss: 11.2512 - acc: 0.1403 - ETA: 23s - loss: 11.2236 - acc: 0.1419 - ETA: 22s - loss: 11.2138 - acc: 0.1419 - ETA: 22s - loss: 11.2213 - acc: 0.1417 - ETA: 22s - loss: 11.2194 - acc: 0.1422 - ETA: 22s - loss: 11.2241 - acc: 0.1416 - ETA: 21s - loss: 11.2160 - acc: 0.1423 - ETA: 21s - loss: 11.2122 - acc: 0.1425 - ETA: 21s - loss: 11.1939 - acc: 0.1432 - ETA: 20s - loss: 11.1834 - acc: 0.1440 - ETA: 20s - loss: 11.1838 - acc: 0.1451 - ETA: 20s - loss: 11.1778 - acc: 0.1455 - ETA: 19s - loss: 11.1719 - acc: 0.1464 - ETA: 19s - loss: 11.1727 - acc: 0.1468 - ETA: 19s - loss: 11.1613 - acc: 0.1474 - ETA: 18s - loss: 11.1388 - acc: 0.1486 - ETA: 18s - loss: 11.1403 - acc: 0.1484 - ETA: 17s - loss: 11.1227 - acc: 0.1492 - ETA: 17s - loss: 11.1289 - acc: 0.1490 - ETA: 16s - loss: 11.1128 - acc: 0.1496 - ETA: 16s - loss: 11.0917 - acc: 0.1502 - ETA: 16s - loss: 11.0812 - acc: 0.1504 - ETA: 15s - loss: 11.0744 - acc: 0.1508 - ETA: 15s - loss: 11.0562 - acc: 0.1517 - ETA: 15s - loss: 11.0493 - acc: 0.1521 - ETA: 14s - loss: 11.0426 - acc: 0.1526 - ETA: 13s - loss: 11.0343 - acc: 0.1535 - ETA: 13s - loss: 11.0355 - acc: 0.1535 - ETA: 13s - loss: 11.0363 - acc: 0.1533 - ETA: 13s - loss: 11.0420 - acc: 0.1530 - ETA: 13s - loss: 11.0359 - acc: 0.1533 - ETA: 12s - loss: 11.0184 - acc: 0.1536 - ETA: 12s - loss: 11.0134 - acc: 0.1538 - ETA: 12s - loss: 11.0061 - acc: 0.1547 - ETA: 11s - loss: 10.9981 - acc: 0.1551 - ETA: 11s - loss: 10.9842 - acc: 0.1565 - ETA: 10s - loss: 10.9650 - acc: 0.1573 - ETA: 10s - loss: 10.9509 - acc: 0.1583 - ETA: 9s - loss: 10.9293 - acc: 0.1589  - ETA: 9s - loss: 10.9393 - acc: 0.1587 - ETA: 9s - loss: 10.9266 - acc: 0.1597 - ETA: 8s - loss: 10.9172 - acc: 0.1612 - ETA: 8s - loss: 10.8956 - acc: 0.1627 - ETA: 7s - loss: 10.8842 - acc: 0.1633 - ETA: 7s - loss: 10.8721 - acc: 0.1639 - ETA: 7s - loss: 10.8713 - acc: 0.1636 - ETA: 7s - loss: 10.8634 - acc: 0.1643 - ETA: 6s - loss: 10.8527 - acc: 0.1652 - ETA: 6s - loss: 10.8480 - acc: 0.1657 - ETA: 5s - loss: 10.8331 - acc: 0.1669 - ETA: 5s - loss: 10.8157 - acc: 0.1681 - ETA: 5s - loss: 10.8045 - acc: 0.1684 - ETA: 4s - loss: 10.8150 - acc: 0.1683 - ETA: 4s - loss: 10.7946 - acc: 0.1694 - ETA: 3s - loss: 10.7871 - acc: 0.1703 - ETA: 3s - loss: 10.7809 - acc: 0.1707 - ETA: 2s - loss: 10.7554 - acc: 0.1723 - ETA: 2s - loss: 10.7601 - acc: 0.1722 - ETA: 2s - loss: 10.7509 - acc: 0.1725 - ETA: 2s - loss: 10.7498 - acc: 0.1729 - ETA: 1s - loss: 10.7351 - acc: 0.1731 - ETA: 1s - loss: 10.7230 - acc: 0.1741 - ETA: 0s - loss: 10.7130 - acc: 0.1751 - ETA: 0s - loss: 10.7161 - acc: 0.1752Epoch 00000: val_loss improved from inf to 9.03461, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 60s - loss: 10.7052 - acc: 0.1756 - val_loss: 9.0346 - val_acc: 0.2838
    Epoch 2/20
    6560/6680 [============================>.] - ETA: 2s - loss: 9.4185 - acc: 0.3000 - ETA: 2s - loss: 8.3044 - acc: 0.3750 - ETA: 2s - loss: 9.1657 - acc: 0.3357 - ETA: 2s - loss: 8.6526 - acc: 0.3525 - ETA: 2s - loss: 8.6996 - acc: 0.3558 - ETA: 2s - loss: 8.7544 - acc: 0.3516 - ETA: 2s - loss: 8.6030 - acc: 0.3566 - ETA: 2s - loss: 8.7185 - acc: 0.3477 - ETA: 2s - loss: 8.7666 - acc: 0.3360 - ETA: 2s - loss: 8.7228 - acc: 0.3375 - ETA: 2s - loss: 8.7406 - acc: 0.3379 - ETA: 2s - loss: 8.7305 - acc: 0.3412 - ETA: 2s - loss: 8.6772 - acc: 0.3419 - ETA: 2s - loss: 8.6607 - acc: 0.3438 - ETA: 2s - loss: 8.6598 - acc: 0.3424 - ETA: 2s - loss: 8.5614 - acc: 0.3495 - ETA: 1s - loss: 8.6005 - acc: 0.3470 - ETA: 1s - loss: 8.5544 - acc: 0.3509 - ETA: 1s - loss: 8.5559 - acc: 0.3500 - ETA: 1s - loss: 8.5281 - acc: 0.3500 - ETA: 1s - loss: 8.5253 - acc: 0.3512 - ETA: 1s - loss: 8.4466 - acc: 0.3563 - ETA: 1s - loss: 8.4269 - acc: 0.3589 - ETA: 1s - loss: 8.4512 - acc: 0.3578 - ETA: 1s - loss: 8.4226 - acc: 0.3597 - ETA: 1s - loss: 8.4037 - acc: 0.3593 - ETA: 1s - loss: 8.3807 - acc: 0.3586 - ETA: 1s - loss: 8.3353 - acc: 0.3595 - ETA: 1s - loss: 8.3019 - acc: 0.3586 - ETA: 1s - loss: 8.2960 - acc: 0.3593 - ETA: 1s - loss: 8.2874 - acc: 0.3594 - ETA: 1s - loss: 8.2476 - acc: 0.3610 - ETA: 0s - loss: 8.2444 - acc: 0.3611 - ETA: 0s - loss: 8.1877 - acc: 0.3635 - ETA: 0s - loss: 8.1571 - acc: 0.3653 - ETA: 0s - loss: 8.1707 - acc: 0.3659 - ETA: 0s - loss: 8.1289 - acc: 0.3692 - ETA: 0s - loss: 8.1014 - acc: 0.3713 - ETA: 0s - loss: 8.0885 - acc: 0.3703 - ETA: 0s - loss: 8.0828 - acc: 0.3707 - ETA: 0s - loss: 8.0970 - acc: 0.3698 - ETA: 0s - loss: 8.0649 - acc: 0.3716 - ETA: 0s - loss: 8.0637 - acc: 0.3707 - ETA: 0s - loss: 8.0572 - acc: 0.3716 - ETA: 0s - loss: 8.0464 - acc: 0.3729 - ETA: 0s - loss: 8.0240 - acc: 0.3745 - ETA: 0s - loss: 7.9931 - acc: 0.3768 - ETA: 0s - loss: 7.9941 - acc: 0.3766 - ETA: 0s - loss: 7.9828 - acc: 0.3770Epoch 00001: val_loss improved from 9.03461 to 7.83952, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 7.9819 - acc: 0.3775 - val_loss: 7.8395 - val_acc: 0.3820
    Epoch 3/20
    6540/6680 [============================>.] - ETA: 2s - loss: 4.3556 - acc: 0.6500 - ETA: 2s - loss: 7.0359 - acc: 0.4667 - ETA: 2s - loss: 7.0886 - acc: 0.4765 - ETA: 2s - loss: 7.3117 - acc: 0.4600 - ETA: 2s - loss: 7.2439 - acc: 0.4656 - ETA: 2s - loss: 7.4731 - acc: 0.4551 - ETA: 2s - loss: 7.3547 - acc: 0.4596 - ETA: 1s - loss: 7.4245 - acc: 0.4583 - ETA: 1s - loss: 7.4699 - acc: 0.4525 - ETA: 1s - loss: 7.4527 - acc: 0.4529 - ETA: 1s - loss: 7.4624 - acc: 0.4547 - ETA: 1s - loss: 7.4146 - acc: 0.4590 - ETA: 1s - loss: 7.3191 - acc: 0.4639 - ETA: 1s - loss: 7.3269 - acc: 0.4593 - ETA: 1s - loss: 7.3762 - acc: 0.4563 - ETA: 1s - loss: 7.3793 - acc: 0.4568 - ETA: 1s - loss: 7.3868 - acc: 0.4572 - ETA: 1s - loss: 7.3540 - acc: 0.4603 - ETA: 1s - loss: 7.3571 - acc: 0.4608 - ETA: 1s - loss: 7.3716 - acc: 0.4613 - ETA: 1s - loss: 7.4036 - acc: 0.4594 - ETA: 1s - loss: 7.3588 - acc: 0.4621 - ETA: 1s - loss: 7.3181 - acc: 0.4652 - ETA: 1s - loss: 7.3154 - acc: 0.4658 - ETA: 1s - loss: 7.3057 - acc: 0.4663 - ETA: 1s - loss: 7.3085 - acc: 0.4661 - ETA: 0s - loss: 7.3184 - acc: 0.4662 - ETA: 0s - loss: 7.3149 - acc: 0.4672 - ETA: 0s - loss: 7.3217 - acc: 0.4675 - ETA: 0s - loss: 7.3051 - acc: 0.4693 - ETA: 0s - loss: 7.3121 - acc: 0.4685 - ETA: 0s - loss: 7.3357 - acc: 0.4662 - ETA: 0s - loss: 7.3602 - acc: 0.4655 - ETA: 0s - loss: 7.3488 - acc: 0.4654 - ETA: 0s - loss: 7.3947 - acc: 0.4621 - ETA: 0s - loss: 7.4048 - acc: 0.4616 - ETA: 0s - loss: 7.4086 - acc: 0.4615 - ETA: 0s - loss: 7.3848 - acc: 0.4624 - ETA: 0s - loss: 7.3719 - acc: 0.4635 - ETA: 0s - loss: 7.3688 - acc: 0.4641 - ETA: 0s - loss: 7.3459 - acc: 0.4644 - ETA: 0s - loss: 7.3305 - acc: 0.4655 - ETA: 0s - loss: 7.3187 - acc: 0.4661 - ETA: 0s - loss: 7.3113 - acc: 0.4663 - ETA: 0s - loss: 7.3075 - acc: 0.4664Epoch 00002: val_loss improved from 7.83952 to 7.70672, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 7.3252 - acc: 0.4653 - val_loss: 7.7067 - val_acc: 0.4036
    Epoch 4/20
    6620/6680 [============================>.] - ETA: 2s - loss: 8.1693 - acc: 0.4500 - ETA: 2s - loss: 6.8777 - acc: 0.5333 - ETA: 2s - loss: 7.1652 - acc: 0.5031 - ETA: 2s - loss: 7.3444 - acc: 0.4935 - ETA: 2s - loss: 7.2706 - acc: 0.4967 - ETA: 2s - loss: 7.2079 - acc: 0.4961 - ETA: 2s - loss: 7.1470 - acc: 0.4989 - ETA: 2s - loss: 7.1092 - acc: 0.5038 - ETA: 1s - loss: 7.0195 - acc: 0.5119 - ETA: 1s - loss: 6.9349 - acc: 0.5144 - ETA: 1s - loss: 6.9643 - acc: 0.5130 - ETA: 1s - loss: 7.0114 - acc: 0.5119 - ETA: 1s - loss: 7.0208 - acc: 0.5085 - ETA: 1s - loss: 7.0253 - acc: 0.5089 - ETA: 1s - loss: 7.0290 - acc: 0.5087 - ETA: 1s - loss: 7.0755 - acc: 0.5059 - ETA: 1s - loss: 7.1118 - acc: 0.5017 - ETA: 1s - loss: 7.1334 - acc: 0.5004 - ETA: 1s - loss: 7.1251 - acc: 0.5011 - ETA: 1s - loss: 7.1189 - acc: 0.5018 - ETA: 1s - loss: 7.1396 - acc: 0.5003 - ETA: 1s - loss: 7.0841 - acc: 0.5029 - ETA: 1s - loss: 7.0429 - acc: 0.5058 - ETA: 1s - loss: 7.0359 - acc: 0.5055 - ETA: 1s - loss: 7.0707 - acc: 0.5045 - ETA: 1s - loss: 7.0758 - acc: 0.5048 - ETA: 0s - loss: 7.1042 - acc: 0.5034 - ETA: 0s - loss: 7.1336 - acc: 0.5008 - ETA: 0s - loss: 7.1267 - acc: 0.5012 - ETA: 0s - loss: 7.0932 - acc: 0.5028 - ETA: 0s - loss: 7.0954 - acc: 0.5022 - ETA: 0s - loss: 7.1120 - acc: 0.5006 - ETA: 0s - loss: 7.0726 - acc: 0.5027 - ETA: 0s - loss: 7.0845 - acc: 0.5027 - ETA: 0s - loss: 7.1241 - acc: 0.5000 - ETA: 0s - loss: 7.1156 - acc: 0.5010 - ETA: 0s - loss: 7.1340 - acc: 0.5004 - ETA: 0s - loss: 7.1460 - acc: 0.5004 - ETA: 0s - loss: 7.1171 - acc: 0.5021 - ETA: 0s - loss: 7.0998 - acc: 0.5033 - ETA: 0s - loss: 7.1497 - acc: 0.4998 - ETA: 0s - loss: 7.1361 - acc: 0.5000 - ETA: 0s - loss: 7.1181 - acc: 0.5006 - ETA: 0s - loss: 7.1050 - acc: 0.5014 - ETA: 0s - loss: 7.1056 - acc: 0.5012 - ETA: 0s - loss: 7.0975 - acc: 0.5021Epoch 00003: val_loss improved from 7.70672 to 7.53867, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 7.1065 - acc: 0.5016 - val_loss: 7.5387 - val_acc: 0.4455
    Epoch 5/20
    6540/6680 [============================>.] - ETA: 2s - loss: 4.2033 - acc: 0.6500 - ETA: 2s - loss: 6.9566 - acc: 0.5000 - ETA: 2s - loss: 6.7364 - acc: 0.5206 - ETA: 2s - loss: 6.9079 - acc: 0.5167 - ETA: 2s - loss: 6.8997 - acc: 0.5194 - ETA: 2s - loss: 6.7474 - acc: 0.5316 - ETA: 2s - loss: 6.7280 - acc: 0.5333 - ETA: 1s - loss: 6.7450 - acc: 0.5356 - ETA: 1s - loss: 6.6944 - acc: 0.5398 - ETA: 1s - loss: 6.7886 - acc: 0.5351 - ETA: 1s - loss: 6.7489 - acc: 0.5365 - ETA: 1s - loss: 6.7577 - acc: 0.5360 - ETA: 1s - loss: 6.7049 - acc: 0.5399 - ETA: 1s - loss: 6.7513 - acc: 0.5354 - ETA: 1s - loss: 6.8512 - acc: 0.5296 - ETA: 1s - loss: 6.8448 - acc: 0.5305 - ETA: 1s - loss: 6.8451 - acc: 0.5295 - ETA: 1s - loss: 6.8865 - acc: 0.5286 - ETA: 1s - loss: 6.9275 - acc: 0.5267 - ETA: 1s - loss: 6.9178 - acc: 0.5268 - ETA: 1s - loss: 6.8680 - acc: 0.5293 - ETA: 1s - loss: 6.9025 - acc: 0.5276 - ETA: 1s - loss: 6.9380 - acc: 0.5261 - ETA: 1s - loss: 6.9248 - acc: 0.5265 - ETA: 1s - loss: 6.9033 - acc: 0.5269 - ETA: 1s - loss: 6.9093 - acc: 0.5267 - ETA: 1s - loss: 6.8964 - acc: 0.5278 - ETA: 1s - loss: 6.8854 - acc: 0.5289 - ETA: 0s - loss: 6.9390 - acc: 0.5248 - ETA: 0s - loss: 6.9440 - acc: 0.5240 - ETA: 0s - loss: 6.9310 - acc: 0.5247 - ETA: 0s - loss: 6.9219 - acc: 0.5257 - ETA: 0s - loss: 6.9464 - acc: 0.5234 - ETA: 0s - loss: 6.9552 - acc: 0.5226 - ETA: 0s - loss: 6.9194 - acc: 0.5253 - ETA: 0s - loss: 6.9022 - acc: 0.5259 - ETA: 0s - loss: 6.9040 - acc: 0.5261 - ETA: 0s - loss: 6.9107 - acc: 0.5258 - ETA: 0s - loss: 6.9083 - acc: 0.5266 - ETA: 0s - loss: 6.9170 - acc: 0.5265 - ETA: 0s - loss: 6.9355 - acc: 0.5257 - ETA: 0s - loss: 6.9426 - acc: 0.5258 - ETA: 0s - loss: 6.9721 - acc: 0.5245 - ETA: 0s - loss: 6.9793 - acc: 0.5244 - ETA: 0s - loss: 6.9671 - acc: 0.5250 - ETA: 0s - loss: 6.9683 - acc: 0.5242Epoch 00004: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 6.9902 - acc: 0.5228 - val_loss: 7.5614 - val_acc: 0.4503
    Epoch 6/20
    6660/6680 [============================>.] - ETA: 2s - loss: 5.7051 - acc: 0.6500 - ETA: 2s - loss: 6.6966 - acc: 0.5500 - ETA: 2s - loss: 7.4227 - acc: 0.4969 - ETA: 2s - loss: 7.3269 - acc: 0.5109 - ETA: 2s - loss: 7.1837 - acc: 0.5250 - ETA: 2s - loss: 7.0661 - acc: 0.5324 - ETA: 2s - loss: 7.1328 - acc: 0.5295 - ETA: 2s - loss: 7.1924 - acc: 0.5265 - ETA: 2s - loss: 7.0749 - acc: 0.5284 - ETA: 1s - loss: 7.0852 - acc: 0.5300 - ETA: 1s - loss: 7.0456 - acc: 0.5347 - ETA: 1s - loss: 7.0869 - acc: 0.5304 - ETA: 1s - loss: 6.9769 - acc: 0.5379 - ETA: 1s - loss: 7.0268 - acc: 0.5362 - ETA: 1s - loss: 6.9855 - acc: 0.5391 - ETA: 1s - loss: 6.9845 - acc: 0.5390 - ETA: 1s - loss: 6.9962 - acc: 0.5359 - ETA: 1s - loss: 6.9896 - acc: 0.5360 - ETA: 1s - loss: 7.0240 - acc: 0.5341 - ETA: 1s - loss: 6.9661 - acc: 0.5379 - ETA: 1s - loss: 6.9840 - acc: 0.5368 - ETA: 1s - loss: 6.9711 - acc: 0.5368 - ETA: 1s - loss: 6.9320 - acc: 0.5386 - ETA: 1s - loss: 6.9026 - acc: 0.5402 - ETA: 1s - loss: 6.9144 - acc: 0.5386 - ETA: 1s - loss: 6.8842 - acc: 0.5410 - ETA: 1s - loss: 6.8781 - acc: 0.5424 - ETA: 0s - loss: 6.8701 - acc: 0.5422 - ETA: 0s - loss: 6.8785 - acc: 0.5415 - ETA: 0s - loss: 6.8819 - acc: 0.5410 - ETA: 0s - loss: 6.8782 - acc: 0.5409 - ETA: 0s - loss: 6.8358 - acc: 0.5430 - ETA: 0s - loss: 6.8200 - acc: 0.5434 - ETA: 0s - loss: 6.8348 - acc: 0.5419 - ETA: 0s - loss: 6.8503 - acc: 0.5408 - ETA: 0s - loss: 6.8759 - acc: 0.5387 - ETA: 0s - loss: 6.8823 - acc: 0.5380 - ETA: 0s - loss: 6.8691 - acc: 0.5391 - ETA: 0s - loss: 6.8789 - acc: 0.5388 - ETA: 0s - loss: 6.8643 - acc: 0.5398 - ETA: 0s - loss: 6.8480 - acc: 0.5409 - ETA: 0s - loss: 6.8674 - acc: 0.5399 - ETA: 0s - loss: 6.8455 - acc: 0.5408 - ETA: 0s - loss: 6.8369 - acc: 0.5415 - ETA: 0s - loss: 6.8478 - acc: 0.5411 - ETA: 0s - loss: 6.8621 - acc: 0.5405 - ETA: 0s - loss: 6.8541 - acc: 0.5408Epoch 00005: val_loss improved from 7.53867 to 7.31742, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.8625 - acc: 0.5404 - val_loss: 7.3174 - val_acc: 0.4659
    Epoch 7/20
    6580/6680 [============================>.] - ETA: 2s - loss: 6.4541 - acc: 0.6000 - ETA: 2s - loss: 6.2565 - acc: 0.5778 - ETA: 2s - loss: 6.6901 - acc: 0.5500 - ETA: 2s - loss: 6.7856 - acc: 0.5438 - ETA: 2s - loss: 6.8102 - acc: 0.5484 - ETA: 2s - loss: 6.8209 - acc: 0.5461 - ETA: 2s - loss: 6.6718 - acc: 0.5543 - ETA: 1s - loss: 6.6422 - acc: 0.5585 - ETA: 1s - loss: 6.7237 - acc: 0.5550 - ETA: 1s - loss: 6.8498 - acc: 0.5478 - ETA: 1s - loss: 6.9137 - acc: 0.5459 - ETA: 1s - loss: 6.9172 - acc: 0.5438 - ETA: 1s - loss: 6.9205 - acc: 0.5443 - ETA: 1s - loss: 6.9825 - acc: 0.5411 - ETA: 1s - loss: 6.9942 - acc: 0.5397 - ETA: 1s - loss: 6.9901 - acc: 0.5399 - ETA: 1s - loss: 6.9870 - acc: 0.5392 - ETA: 1s - loss: 6.9326 - acc: 0.5411 - ETA: 1s - loss: 6.9260 - acc: 0.5408 - ETA: 1s - loss: 6.8954 - acc: 0.5427 - ETA: 1s - loss: 6.8823 - acc: 0.5444 - ETA: 1s - loss: 6.8818 - acc: 0.5437 - ETA: 1s - loss: 6.8972 - acc: 0.5434 - ETA: 1s - loss: 6.9211 - acc: 0.5418 - ETA: 1s - loss: 6.8641 - acc: 0.5448 - ETA: 1s - loss: 6.8753 - acc: 0.5453 - ETA: 1s - loss: 6.8678 - acc: 0.5455 - ETA: 1s - loss: 6.8332 - acc: 0.5477 - ETA: 0s - loss: 6.8497 - acc: 0.5468 - ETA: 0s - loss: 6.8548 - acc: 0.5471 - ETA: 0s - loss: 6.8307 - acc: 0.5495 - ETA: 0s - loss: 6.8376 - acc: 0.5491 - ETA: 0s - loss: 6.8418 - acc: 0.5489 - ETA: 0s - loss: 6.8260 - acc: 0.5502 - ETA: 0s - loss: 6.8383 - acc: 0.5494 - ETA: 0s - loss: 6.8303 - acc: 0.5500 - ETA: 0s - loss: 6.8335 - acc: 0.5500 - ETA: 0s - loss: 6.8071 - acc: 0.5509 - ETA: 0s - loss: 6.7855 - acc: 0.5520 - ETA: 0s - loss: 6.7748 - acc: 0.5527 - ETA: 0s - loss: 6.7856 - acc: 0.5524 - ETA: 0s - loss: 6.7606 - acc: 0.5540 - ETA: 0s - loss: 6.7512 - acc: 0.5544 - ETA: 0s - loss: 6.7615 - acc: 0.5542 - ETA: 0s - loss: 6.7766 - acc: 0.5533 - ETA: 0s - loss: 6.7694 - acc: 0.5538Epoch 00006: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 6.7717 - acc: 0.5534 - val_loss: 7.3525 - val_acc: 0.4659
    Epoch 8/20
    6580/6680 [============================>.] - ETA: 2s - loss: 8.9788 - acc: 0.4000 - ETA: 2s - loss: 7.0044 - acc: 0.5500 - ETA: 2s - loss: 6.7569 - acc: 0.5667 - ETA: 2s - loss: 6.7802 - acc: 0.5659 - ETA: 2s - loss: 6.4811 - acc: 0.5862 - ETA: 2s - loss: 6.3200 - acc: 0.5919 - ETA: 2s - loss: 6.2149 - acc: 0.5956 - ETA: 2s - loss: 6.2211 - acc: 0.5952 - ETA: 1s - loss: 6.1641 - acc: 0.6000 - ETA: 1s - loss: 6.3187 - acc: 0.5909 - ETA: 1s - loss: 6.4306 - acc: 0.5836 - ETA: 1s - loss: 6.3747 - acc: 0.5869 - ETA: 1s - loss: 6.3350 - acc: 0.5897 - ETA: 1s - loss: 6.4041 - acc: 0.5830 - ETA: 1s - loss: 6.3579 - acc: 0.5842 - ETA: 1s - loss: 6.4206 - acc: 0.5806 - ETA: 1s - loss: 6.4952 - acc: 0.5761 - ETA: 1s - loss: 6.5117 - acc: 0.5754 - ETA: 1s - loss: 6.5418 - acc: 0.5733 - ETA: 1s - loss: 6.5917 - acc: 0.5706 - ETA: 1s - loss: 6.6733 - acc: 0.5647 - ETA: 1s - loss: 6.7333 - acc: 0.5617 - ETA: 1s - loss: 6.7465 - acc: 0.5596 - ETA: 1s - loss: 6.7432 - acc: 0.5598 - ETA: 1s - loss: 6.7760 - acc: 0.5582 - ETA: 1s - loss: 6.7743 - acc: 0.5587 - ETA: 1s - loss: 6.7859 - acc: 0.5576 - ETA: 1s - loss: 6.7844 - acc: 0.5578 - ETA: 0s - loss: 6.7377 - acc: 0.5611 - ETA: 0s - loss: 6.6998 - acc: 0.5626 - ETA: 0s - loss: 6.6923 - acc: 0.5636 - ETA: 0s - loss: 6.6652 - acc: 0.5655 - ETA: 0s - loss: 6.6408 - acc: 0.5675 - ETA: 0s - loss: 6.6360 - acc: 0.5681 - ETA: 0s - loss: 6.6710 - acc: 0.5660 - ETA: 0s - loss: 6.6863 - acc: 0.5650 - ETA: 0s - loss: 6.7347 - acc: 0.5623 - ETA: 0s - loss: 6.7639 - acc: 0.5598 - ETA: 0s - loss: 6.7610 - acc: 0.5599 - ETA: 0s - loss: 6.7684 - acc: 0.5593 - ETA: 0s - loss: 6.7870 - acc: 0.5584 - ETA: 0s - loss: 6.7868 - acc: 0.5582 - ETA: 0s - loss: 6.7705 - acc: 0.5593 - ETA: 0s - loss: 6.7622 - acc: 0.5602 - ETA: 0s - loss: 6.7287 - acc: 0.5622 - ETA: 0s - loss: 6.7278 - acc: 0.5618 - ETA: 0s - loss: 6.7292 - acc: 0.5616Epoch 00007: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 6.7333 - acc: 0.5614 - val_loss: 7.3797 - val_acc: 0.4563
    Epoch 9/20
    6620/6680 [============================>.] - ETA: 2s - loss: 9.6767 - acc: 0.4000 - ETA: 2s - loss: 7.2570 - acc: 0.5500 - ETA: 2s - loss: 6.7141 - acc: 0.5767 - ETA: 2s - loss: 6.8227 - acc: 0.5705 - ETA: 2s - loss: 6.9586 - acc: 0.5638 - ETA: 2s - loss: 6.6818 - acc: 0.5806 - ETA: 2s - loss: 6.8920 - acc: 0.5674 - ETA: 2s - loss: 6.9281 - acc: 0.5650 - ETA: 2s - loss: 6.8284 - acc: 0.5719 - ETA: 2s - loss: 6.8533 - acc: 0.5695 - ETA: 1s - loss: 6.8554 - acc: 0.5683 - ETA: 1s - loss: 6.8120 - acc: 0.5705 - ETA: 1s - loss: 6.8423 - acc: 0.5682 - ETA: 1s - loss: 6.7983 - acc: 0.5701 - ETA: 1s - loss: 6.7701 - acc: 0.5717 - ETA: 1s - loss: 6.7510 - acc: 0.5731 - ETA: 1s - loss: 6.7062 - acc: 0.5752 - ETA: 1s - loss: 6.7409 - acc: 0.5725 - ETA: 1s - loss: 6.7073 - acc: 0.5742 - ETA: 1s - loss: 6.7743 - acc: 0.5704 - ETA: 1s - loss: 6.7982 - acc: 0.5694 - ETA: 1s - loss: 6.8117 - acc: 0.5681 - ETA: 1s - loss: 6.8125 - acc: 0.5679 - ETA: 1s - loss: 6.7903 - acc: 0.5690 - ETA: 1s - loss: 6.8080 - acc: 0.5676 - ETA: 1s - loss: 6.7838 - acc: 0.5686 - ETA: 1s - loss: 6.7559 - acc: 0.5699 - ETA: 1s - loss: 6.7468 - acc: 0.5704 - ETA: 0s - loss: 6.7417 - acc: 0.5698 - ETA: 0s - loss: 6.7341 - acc: 0.5703 - ETA: 0s - loss: 6.7294 - acc: 0.5698 - ETA: 0s - loss: 6.7223 - acc: 0.5704 - ETA: 0s - loss: 6.7244 - acc: 0.5701 - ETA: 0s - loss: 6.7146 - acc: 0.5707 - ETA: 0s - loss: 6.7112 - acc: 0.5709 - ETA: 0s - loss: 6.6769 - acc: 0.5731 - ETA: 0s - loss: 6.7067 - acc: 0.5713 - ETA: 0s - loss: 6.7159 - acc: 0.5709 - ETA: 0s - loss: 6.6982 - acc: 0.5718 - ETA: 0s - loss: 6.6831 - acc: 0.5725 - ETA: 0s - loss: 6.6807 - acc: 0.5728 - ETA: 0s - loss: 6.6727 - acc: 0.5736 - ETA: 0s - loss: 6.7125 - acc: 0.5713 - ETA: 0s - loss: 6.7122 - acc: 0.5715 - ETA: 0s - loss: 6.7023 - acc: 0.5719 - ETA: 0s - loss: 6.7114 - acc: 0.5711 - ETA: 0s - loss: 6.7092 - acc: 0.5713Epoch 00008: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 6.7007 - acc: 0.5716 - val_loss: 7.3487 - val_acc: 0.4707
    Epoch 10/20
    6560/6680 [============================>.] - ETA: 2s - loss: 6.4476 - acc: 0.6000 - ETA: 2s - loss: 6.6616 - acc: 0.5813 - ETA: 2s - loss: 6.4150 - acc: 0.5967 - ETA: 2s - loss: 6.5021 - acc: 0.5932 - ETA: 2s - loss: 6.6949 - acc: 0.5776 - ETA: 2s - loss: 6.5855 - acc: 0.5847 - ETA: 2s - loss: 6.5225 - acc: 0.5849 - ETA: 2s - loss: 6.7222 - acc: 0.5740 - ETA: 2s - loss: 6.8335 - acc: 0.5667 - ETA: 1s - loss: 6.7915 - acc: 0.5703 - ETA: 1s - loss: 6.5900 - acc: 0.5824 - ETA: 1s - loss: 6.5542 - acc: 0.5833 - ETA: 1s - loss: 6.5954 - acc: 0.5812 - ETA: 1s - loss: 6.7304 - acc: 0.5728 - ETA: 1s - loss: 6.8094 - acc: 0.5685 - ETA: 1s - loss: 6.8336 - acc: 0.5668 - ETA: 1s - loss: 6.8053 - acc: 0.5684 - ETA: 1s - loss: 6.7795 - acc: 0.5698 - ETA: 1s - loss: 6.7689 - acc: 0.5703 - ETA: 1s - loss: 6.7465 - acc: 0.5719 - ETA: 1s - loss: 6.7385 - acc: 0.5725 - ETA: 1s - loss: 6.7203 - acc: 0.5738 - ETA: 1s - loss: 6.7159 - acc: 0.5734 - ETA: 1s - loss: 6.6974 - acc: 0.5747 - ETA: 1s - loss: 6.7489 - acc: 0.5713 - ETA: 1s - loss: 6.7411 - acc: 0.5707 - ETA: 1s - loss: 6.7241 - acc: 0.5714 - ETA: 1s - loss: 6.6903 - acc: 0.5728 - ETA: 0s - loss: 6.7212 - acc: 0.5714 - ETA: 0s - loss: 6.7694 - acc: 0.5682 - ETA: 0s - loss: 6.7618 - acc: 0.5687 - ETA: 0s - loss: 6.7474 - acc: 0.5692 - ETA: 0s - loss: 6.7309 - acc: 0.5704 - ETA: 0s - loss: 6.7093 - acc: 0.5716 - ETA: 0s - loss: 6.6958 - acc: 0.5723 - ETA: 0s - loss: 6.6975 - acc: 0.5721 - ETA: 0s - loss: 6.7106 - acc: 0.5708 - ETA: 0s - loss: 6.7341 - acc: 0.5695 - ETA: 0s - loss: 6.7122 - acc: 0.5703 - ETA: 0s - loss: 6.7007 - acc: 0.5701 - ETA: 0s - loss: 6.6758 - acc: 0.5713 - ETA: 0s - loss: 6.6831 - acc: 0.5707 - ETA: 0s - loss: 6.6638 - acc: 0.5715 - ETA: 0s - loss: 6.6568 - acc: 0.5713 - ETA: 0s - loss: 6.6471 - acc: 0.5718 - ETA: 0s - loss: 6.6597 - acc: 0.5712Epoch 00009: val_loss improved from 7.31742 to 7.28554, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.6718 - acc: 0.5704 - val_loss: 7.2855 - val_acc: 0.4707
    Epoch 11/20
    6660/6680 [============================>.] - ETA: 2s - loss: 4.8530 - acc: 0.7000 - ETA: 2s - loss: 7.1603 - acc: 0.5312 - ETA: 2s - loss: 7.1609 - acc: 0.5367 - ETA: 2s - loss: 6.9731 - acc: 0.5523 - ETA: 2s - loss: 6.9980 - acc: 0.5517 - ETA: 2s - loss: 6.8067 - acc: 0.5639 - ETA: 2s - loss: 6.9005 - acc: 0.5593 - ETA: 2s - loss: 6.8004 - acc: 0.5630 - ETA: 2s - loss: 6.8031 - acc: 0.5623 - ETA: 1s - loss: 6.7780 - acc: 0.5648 - ETA: 1s - loss: 6.7581 - acc: 0.5648 - ETA: 1s - loss: 6.6527 - acc: 0.5724 - ETA: 1s - loss: 6.6501 - acc: 0.5724 - ETA: 1s - loss: 6.6484 - acc: 0.5734 - ETA: 1s - loss: 6.5668 - acc: 0.5768 - ETA: 1s - loss: 6.5546 - acc: 0.5774 - ETA: 1s - loss: 6.5610 - acc: 0.5770 - ETA: 1s - loss: 6.5684 - acc: 0.5767 - ETA: 1s - loss: 6.5252 - acc: 0.5783 - ETA: 1s - loss: 6.5543 - acc: 0.5761 - ETA: 1s - loss: 6.5727 - acc: 0.5748 - ETA: 1s - loss: 6.5272 - acc: 0.5774 - ETA: 1s - loss: 6.5589 - acc: 0.5758 - ETA: 1s - loss: 6.5679 - acc: 0.5753 - ETA: 1s - loss: 6.6078 - acc: 0.5729 - ETA: 1s - loss: 6.5657 - acc: 0.5760 - ETA: 1s - loss: 6.5849 - acc: 0.5746 - ETA: 1s - loss: 6.5694 - acc: 0.5751 - ETA: 0s - loss: 6.5581 - acc: 0.5761 - ETA: 0s - loss: 6.6041 - acc: 0.5734 - ETA: 0s - loss: 6.6149 - acc: 0.5729 - ETA: 0s - loss: 6.6077 - acc: 0.5735 - ETA: 0s - loss: 6.5821 - acc: 0.5753 - ETA: 0s - loss: 6.5709 - acc: 0.5757 - ETA: 0s - loss: 6.5830 - acc: 0.5750 - ETA: 0s - loss: 6.5846 - acc: 0.5750 - ETA: 0s - loss: 6.5887 - acc: 0.5750 - ETA: 0s - loss: 6.5973 - acc: 0.5746 - ETA: 0s - loss: 6.6213 - acc: 0.5724 - ETA: 0s - loss: 6.5808 - acc: 0.5743 - ETA: 0s - loss: 6.5782 - acc: 0.5737 - ETA: 0s - loss: 6.5494 - acc: 0.5752 - ETA: 0s - loss: 6.5503 - acc: 0.5752 - ETA: 0s - loss: 6.5534 - acc: 0.5750 - ETA: 0s - loss: 6.5562 - acc: 0.5745 - ETA: 0s - loss: 6.5590 - acc: 0.5747 - ETA: 0s - loss: 6.5476 - acc: 0.5749Epoch 00010: val_loss improved from 7.28554 to 7.19207, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.5329 - acc: 0.5759 - val_loss: 7.1921 - val_acc: 0.4719
    Epoch 12/20
    6560/6680 [============================>.] - ETA: 2s - loss: 7.2533 - acc: 0.5500 - ETA: 2s - loss: 8.1674 - acc: 0.4812 - ETA: 2s - loss: 7.2200 - acc: 0.5433 - ETA: 2s - loss: 7.0332 - acc: 0.5568 - ETA: 2s - loss: 6.7857 - acc: 0.5724 - ETA: 2s - loss: 6.7798 - acc: 0.5722 - ETA: 2s - loss: 6.5746 - acc: 0.5826 - ETA: 2s - loss: 6.6470 - acc: 0.5780 - ETA: 2s - loss: 6.6205 - acc: 0.5807 - ETA: 2s - loss: 6.5691 - acc: 0.5836 - ETA: 1s - loss: 6.5269 - acc: 0.5859 - ETA: 1s - loss: 6.5525 - acc: 0.5808 - ETA: 1s - loss: 6.5366 - acc: 0.5818 - ETA: 1s - loss: 6.5303 - acc: 0.5810 - ETA: 1s - loss: 6.5181 - acc: 0.5818 - ETA: 1s - loss: 6.4699 - acc: 0.5854 - ETA: 1s - loss: 6.4850 - acc: 0.5841 - ETA: 1s - loss: 6.5629 - acc: 0.5788 - ETA: 1s - loss: 6.5367 - acc: 0.5811 - ETA: 1s - loss: 6.5457 - acc: 0.5802 - ETA: 1s - loss: 6.5184 - acc: 0.5826 - ETA: 1s - loss: 6.4770 - acc: 0.5858 - ETA: 1s - loss: 6.4267 - acc: 0.5884 - ETA: 1s - loss: 6.4195 - acc: 0.5893 - ETA: 1s - loss: 6.4038 - acc: 0.5903 - ETA: 1s - loss: 6.4216 - acc: 0.5893 - ETA: 1s - loss: 6.4178 - acc: 0.5894 - ETA: 1s - loss: 6.4035 - acc: 0.5901 - ETA: 1s - loss: 6.4215 - acc: 0.5894 - ETA: 0s - loss: 6.4684 - acc: 0.5861 - ETA: 0s - loss: 6.4906 - acc: 0.5844 - ETA: 0s - loss: 6.4979 - acc: 0.5839 - ETA: 0s - loss: 6.4943 - acc: 0.5836 - ETA: 0s - loss: 6.4803 - acc: 0.5839 - ETA: 0s - loss: 6.4807 - acc: 0.5836 - ETA: 0s - loss: 6.5044 - acc: 0.5819 - ETA: 0s - loss: 6.5078 - acc: 0.5820 - ETA: 0s - loss: 6.4963 - acc: 0.5828 - ETA: 0s - loss: 6.4692 - acc: 0.5845 - ETA: 0s - loss: 6.4475 - acc: 0.5860 - ETA: 0s - loss: 6.4530 - acc: 0.5857 - ETA: 0s - loss: 6.4548 - acc: 0.5854 - ETA: 0s - loss: 6.4249 - acc: 0.5871 - ETA: 0s - loss: 6.4194 - acc: 0.5874 - ETA: 0s - loss: 6.4140 - acc: 0.5872 - ETA: 0s - loss: 6.3838 - acc: 0.5890Epoch 00011: val_loss improved from 7.19207 to 7.14064, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.3888 - acc: 0.5886 - val_loss: 7.1406 - val_acc: 0.4623
    Epoch 13/20
    6540/6680 [============================>.] - ETA: 2s - loss: 6.4485 - acc: 0.6000 - ETA: 2s - loss: 7.2638 - acc: 0.5500 - ETA: 2s - loss: 7.0688 - acc: 0.5500 - ETA: 2s - loss: 6.7135 - acc: 0.5659 - ETA: 2s - loss: 6.5108 - acc: 0.5724 - ETA: 2s - loss: 6.4742 - acc: 0.5778 - ETA: 2s - loss: 6.4875 - acc: 0.5779 - ETA: 2s - loss: 6.4018 - acc: 0.5820 - ETA: 2s - loss: 6.3713 - acc: 0.5833 - ETA: 1s - loss: 6.3204 - acc: 0.5859 - ETA: 1s - loss: 6.2626 - acc: 0.5894 - ETA: 1s - loss: 6.2010 - acc: 0.5929 - ETA: 1s - loss: 6.1842 - acc: 0.5924 - ETA: 1s - loss: 6.2301 - acc: 0.5875 - ETA: 1s - loss: 6.2299 - acc: 0.5879 - ETA: 1s - loss: 6.1705 - acc: 0.5915 - ETA: 1s - loss: 6.2412 - acc: 0.5867 - ETA: 1s - loss: 6.2400 - acc: 0.5854 - ETA: 1s - loss: 6.2596 - acc: 0.5854 - ETA: 1s - loss: 6.2511 - acc: 0.5866 - ETA: 1s - loss: 6.2301 - acc: 0.5887 - ETA: 1s - loss: 6.2270 - acc: 0.5892 - ETA: 1s - loss: 6.1988 - acc: 0.5913 - ETA: 1s - loss: 6.2259 - acc: 0.5901 - ETA: 1s - loss: 6.2148 - acc: 0.5902 - ETA: 1s - loss: 6.2369 - acc: 0.5889 - ETA: 1s - loss: 6.2294 - acc: 0.5891 - ETA: 1s - loss: 6.2250 - acc: 0.5895 - ETA: 1s - loss: 6.2366 - acc: 0.5886 - ETA: 0s - loss: 6.2398 - acc: 0.5887 - ETA: 0s - loss: 6.2668 - acc: 0.5877 - ETA: 0s - loss: 6.2475 - acc: 0.5892 - ETA: 0s - loss: 6.2617 - acc: 0.5882 - ETA: 0s - loss: 6.2595 - acc: 0.5888 - ETA: 0s - loss: 6.2535 - acc: 0.5889 - ETA: 0s - loss: 6.2527 - acc: 0.5892 - ETA: 0s - loss: 6.2648 - acc: 0.5883 - ETA: 0s - loss: 6.2151 - acc: 0.5915 - ETA: 0s - loss: 6.2328 - acc: 0.5904 - ETA: 0s - loss: 6.2152 - acc: 0.5918 - ETA: 0s - loss: 6.2019 - acc: 0.5929 - ETA: 0s - loss: 6.2234 - acc: 0.5913 - ETA: 0s - loss: 6.1938 - acc: 0.5936 - ETA: 0s - loss: 6.2053 - acc: 0.5926 - ETA: 0s - loss: 6.1906 - acc: 0.5939 - ETA: 0s - loss: 6.1780 - acc: 0.5950 - ETA: 0s - loss: 6.1864 - acc: 0.5945Epoch 00012: val_loss improved from 7.14064 to 6.96708, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.2003 - acc: 0.5934 - val_loss: 6.9671 - val_acc: 0.4922
    Epoch 14/20
    6580/6680 [============================>.] - ETA: 2s - loss: 4.0318 - acc: 0.7500 - ETA: 2s - loss: 5.6627 - acc: 0.6389 - ETA: 2s - loss: 6.0738 - acc: 0.6147 - ETA: 2s - loss: 6.2663 - acc: 0.5979 - ETA: 2s - loss: 6.2076 - acc: 0.6016 - ETA: 2s - loss: 6.1520 - acc: 0.6038 - ETA: 2s - loss: 6.1807 - acc: 0.6021 - ETA: 1s - loss: 6.1734 - acc: 0.6037 - ETA: 1s - loss: 6.2135 - acc: 0.6016 - ETA: 1s - loss: 6.1295 - acc: 0.6065 - ETA: 1s - loss: 6.0458 - acc: 0.6105 - ETA: 1s - loss: 6.1729 - acc: 0.6030 - ETA: 1s - loss: 6.1594 - acc: 0.6050 - ETA: 1s - loss: 6.1656 - acc: 0.6046 - ETA: 1s - loss: 6.1203 - acc: 0.6067 - ETA: 1s - loss: 6.1290 - acc: 0.6059 - ETA: 1s - loss: 6.1366 - acc: 0.6051 - ETA: 1s - loss: 6.0973 - acc: 0.6076 - ETA: 1s - loss: 6.1346 - acc: 0.6061 - ETA: 1s - loss: 6.1299 - acc: 0.6065 - ETA: 1s - loss: 6.1396 - acc: 0.6055 - ETA: 1s - loss: 6.1195 - acc: 0.6069 - ETA: 1s - loss: 6.0895 - acc: 0.6088 - ETA: 1s - loss: 6.0725 - acc: 0.6090 - ETA: 1s - loss: 6.0662 - acc: 0.6092 - ETA: 1s - loss: 6.0370 - acc: 0.6115 - ETA: 1s - loss: 6.0281 - acc: 0.6124 - ETA: 1s - loss: 6.0449 - acc: 0.6112 - ETA: 0s - loss: 6.0813 - acc: 0.6089 - ETA: 0s - loss: 6.0907 - acc: 0.6086 - ETA: 0s - loss: 6.0765 - acc: 0.6099 - ETA: 0s - loss: 6.0727 - acc: 0.6096 - ETA: 0s - loss: 6.0884 - acc: 0.6087 - ETA: 0s - loss: 6.1221 - acc: 0.6063 - ETA: 0s - loss: 6.1125 - acc: 0.6065 - ETA: 0s - loss: 6.1444 - acc: 0.6050 - ETA: 0s - loss: 6.1175 - acc: 0.6062 - ETA: 0s - loss: 6.1136 - acc: 0.6062 - ETA: 0s - loss: 6.1154 - acc: 0.6060 - ETA: 0s - loss: 6.1084 - acc: 0.6064 - ETA: 0s - loss: 6.0879 - acc: 0.6075 - ETA: 0s - loss: 6.0868 - acc: 0.6077 - ETA: 0s - loss: 6.0725 - acc: 0.6085 - ETA: 0s - loss: 6.0796 - acc: 0.6071 - ETA: 0s - loss: 6.0605 - acc: 0.6086 - ETA: 0s - loss: 6.0623 - acc: 0.6085 - ETA: 0s - loss: 6.0799 - acc: 0.6073Epoch 00013: val_loss improved from 6.96708 to 6.90283, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 6.0955 - acc: 0.6063 - val_loss: 6.9028 - val_acc: 0.4946
    Epoch 15/20
    6640/6680 [============================>.] - ETA: 2s - loss: 6.5341 - acc: 0.5500 - ETA: 2s - loss: 6.7830 - acc: 0.5625 - ETA: 2s - loss: 6.4974 - acc: 0.5767 - ETA: 2s - loss: 6.1607 - acc: 0.6000 - ETA: 2s - loss: 6.0086 - acc: 0.6086 - ETA: 2s - loss: 5.9323 - acc: 0.6139 - ETA: 2s - loss: 5.8319 - acc: 0.6221 - ETA: 2s - loss: 5.9466 - acc: 0.6160 - ETA: 2s - loss: 5.9876 - acc: 0.6132 - ETA: 2s - loss: 5.9460 - acc: 0.6164 - ETA: 1s - loss: 5.9303 - acc: 0.6176 - ETA: 1s - loss: 5.8492 - acc: 0.6209 - ETA: 1s - loss: 5.9007 - acc: 0.6172 - ETA: 1s - loss: 5.8693 - acc: 0.6189 - ETA: 1s - loss: 5.8646 - acc: 0.6194 - ETA: 1s - loss: 5.8897 - acc: 0.6182 - ETA: 1s - loss: 5.8242 - acc: 0.6220 - ETA: 1s - loss: 5.9051 - acc: 0.6176 - ETA: 1s - loss: 5.9373 - acc: 0.6163 - ETA: 1s - loss: 5.9325 - acc: 0.6154 - ETA: 1s - loss: 5.9504 - acc: 0.6136 - ETA: 1s - loss: 5.9162 - acc: 0.6165 - ETA: 1s - loss: 5.9279 - acc: 0.6151 - ETA: 1s - loss: 5.9604 - acc: 0.6129 - ETA: 1s - loss: 5.9647 - acc: 0.6127 - ETA: 1s - loss: 5.9629 - acc: 0.6124 - ETA: 1s - loss: 5.9857 - acc: 0.6112 - ETA: 0s - loss: 6.0692 - acc: 0.6060 - ETA: 0s - loss: 6.0713 - acc: 0.6061 - ETA: 0s - loss: 6.0625 - acc: 0.6063 - ETA: 0s - loss: 6.0309 - acc: 0.6084 - ETA: 0s - loss: 6.0380 - acc: 0.6077 - ETA: 0s - loss: 6.0347 - acc: 0.6081 - ETA: 0s - loss: 6.0594 - acc: 0.6064 - ETA: 0s - loss: 6.0683 - acc: 0.6058 - ETA: 0s - loss: 6.0563 - acc: 0.6067 - ETA: 0s - loss: 6.0331 - acc: 0.6086 - ETA: 0s - loss: 6.0343 - acc: 0.6086 - ETA: 0s - loss: 6.0292 - acc: 0.6089 - ETA: 0s - loss: 5.9970 - acc: 0.6104 - ETA: 0s - loss: 5.9836 - acc: 0.6116 - ETA: 0s - loss: 5.9653 - acc: 0.6121 - ETA: 0s - loss: 5.9798 - acc: 0.6113 - ETA: 0s - loss: 5.9535 - acc: 0.6132 - ETA: 0s - loss: 5.9545 - acc: 0.6132 - ETA: 0s - loss: 5.9593 - acc: 0.6129 - ETA: 0s - loss: 5.9504 - acc: 0.6139Epoch 00014: val_loss improved from 6.90283 to 6.77371, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 5.9538 - acc: 0.6135 - val_loss: 6.7737 - val_acc: 0.4934
    Epoch 16/20
    6600/6680 [============================>.] - ETA: 2s - loss: 4.8368 - acc: 0.7000 - ETA: 2s - loss: 5.9563 - acc: 0.6250 - ETA: 2s - loss: 5.4972 - acc: 0.6500 - ETA: 2s - loss: 5.7275 - acc: 0.6386 - ETA: 2s - loss: 5.7081 - acc: 0.6414 - ETA: 2s - loss: 5.6594 - acc: 0.6389 - ETA: 2s - loss: 5.4548 - acc: 0.6512 - ETA: 2s - loss: 5.5913 - acc: 0.6420 - ETA: 2s - loss: 5.7438 - acc: 0.6333 - ETA: 2s - loss: 5.7712 - acc: 0.6328 - ETA: 1s - loss: 5.8923 - acc: 0.6239 - ETA: 1s - loss: 5.8970 - acc: 0.6224 - ETA: 1s - loss: 5.9446 - acc: 0.6188 - ETA: 1s - loss: 5.9168 - acc: 0.6201 - ETA: 1s - loss: 5.9096 - acc: 0.6197 - ETA: 1s - loss: 5.9154 - acc: 0.6193 - ETA: 1s - loss: 5.9380 - acc: 0.6177 - ETA: 1s - loss: 5.9020 - acc: 0.6200 - ETA: 1s - loss: 5.8457 - acc: 0.6236 - ETA: 1s - loss: 5.9021 - acc: 0.6205 - ETA: 1s - loss: 5.8932 - acc: 0.6213 - ETA: 1s - loss: 5.9121 - acc: 0.6203 - ETA: 1s - loss: 5.9238 - acc: 0.6194 - ETA: 1s - loss: 5.9123 - acc: 0.6204 - ETA: 1s - loss: 5.9412 - acc: 0.6186 - ETA: 1s - loss: 5.9054 - acc: 0.6210 - ETA: 1s - loss: 5.9116 - acc: 0.6208 - ETA: 1s - loss: 5.9065 - acc: 0.6216 - ETA: 1s - loss: 5.9390 - acc: 0.6195 - ETA: 0s - loss: 5.9142 - acc: 0.6211 - ETA: 0s - loss: 5.8684 - acc: 0.6239 - ETA: 0s - loss: 5.8679 - acc: 0.6241 - ETA: 0s - loss: 5.8825 - acc: 0.6233 - ETA: 0s - loss: 5.8983 - acc: 0.6224 - ETA: 0s - loss: 5.8877 - acc: 0.6225 - ETA: 0s - loss: 5.9102 - acc: 0.6211 - ETA: 0s - loss: 5.9033 - acc: 0.6217 - ETA: 0s - loss: 5.9147 - acc: 0.6210 - ETA: 0s - loss: 5.9031 - acc: 0.6213 - ETA: 0s - loss: 5.9128 - acc: 0.6206 - ETA: 0s - loss: 5.9084 - acc: 0.6211 - ETA: 0s - loss: 5.9077 - acc: 0.6212 - ETA: 0s - loss: 5.9046 - acc: 0.6215 - ETA: 0s - loss: 5.8806 - acc: 0.6226 - ETA: 0s - loss: 5.8795 - acc: 0.6224 - ETA: 0s - loss: 5.8772 - acc: 0.6222 - ETA: 0s - loss: 5.8633 - acc: 0.6229Epoch 00015: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 5.8557 - acc: 0.6231 - val_loss: 6.7866 - val_acc: 0.4802
    Epoch 17/20
    6540/6680 [============================>.] - ETA: 2s - loss: 4.8449 - acc: 0.7000 - ETA: 2s - loss: 5.2793 - acc: 0.6611 - ETA: 2s - loss: 5.7901 - acc: 0.6312 - ETA: 2s - loss: 5.6799 - acc: 0.6396 - ETA: 2s - loss: 5.7085 - acc: 0.6359 - ETA: 2s - loss: 5.8191 - acc: 0.6300 - ETA: 1s - loss: 5.8822 - acc: 0.6255 - ETA: 1s - loss: 5.8433 - acc: 0.6259 - ETA: 1s - loss: 5.8998 - acc: 0.6238 - ETA: 1s - loss: 5.9920 - acc: 0.6191 - ETA: 1s - loss: 5.8876 - acc: 0.6260 - ETA: 1s - loss: 5.8577 - acc: 0.6280 - ETA: 1s - loss: 5.8453 - acc: 0.6287 - ETA: 1s - loss: 5.7454 - acc: 0.6344 - ETA: 1s - loss: 5.6916 - acc: 0.6383 - ETA: 1s - loss: 5.7329 - acc: 0.6364 - ETA: 1s - loss: 5.8036 - acc: 0.6325 - ETA: 1s - loss: 5.8028 - acc: 0.6310 - ETA: 1s - loss: 5.8865 - acc: 0.6263 - ETA: 1s - loss: 5.8458 - acc: 0.6290 - ETA: 1s - loss: 5.8696 - acc: 0.6276 - ETA: 1s - loss: 5.9230 - acc: 0.6247 - ETA: 1s - loss: 5.9369 - acc: 0.6239 - ETA: 1s - loss: 5.9498 - acc: 0.6229 - ETA: 1s - loss: 5.9562 - acc: 0.6228 - ETA: 1s - loss: 5.9173 - acc: 0.6256 - ETA: 1s - loss: 5.9339 - acc: 0.6241 - ETA: 1s - loss: 5.9414 - acc: 0.6237 - ETA: 0s - loss: 5.9275 - acc: 0.6244 - ETA: 0s - loss: 5.9219 - acc: 0.6250 - ETA: 0s - loss: 5.9434 - acc: 0.6235 - ETA: 0s - loss: 5.9414 - acc: 0.6236 - ETA: 0s - loss: 5.9556 - acc: 0.6225 - ETA: 0s - loss: 5.9560 - acc: 0.6222 - ETA: 0s - loss: 5.9423 - acc: 0.6228 - ETA: 0s - loss: 5.9256 - acc: 0.6240 - ETA: 0s - loss: 5.8999 - acc: 0.6257 - ETA: 0s - loss: 5.8597 - acc: 0.6280 - ETA: 0s - loss: 5.8675 - acc: 0.6273 - ETA: 0s - loss: 5.8408 - acc: 0.6288 - ETA: 0s - loss: 5.8254 - acc: 0.6296 - ETA: 0s - loss: 5.8409 - acc: 0.6288 - ETA: 0s - loss: 5.8177 - acc: 0.6303 - ETA: 0s - loss: 5.8359 - acc: 0.6291 - ETA: 0s - loss: 5.8104 - acc: 0.6302 - ETA: 0s - loss: 5.8068 - acc: 0.6300 - ETA: 0s - loss: 5.8113 - acc: 0.6298Epoch 00016: val_loss improved from 6.77371 to 6.73259, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 5.8259 - acc: 0.6290 - val_loss: 6.7326 - val_acc: 0.5018
    Epoch 18/20
    6560/6680 [============================>.] - ETA: 2s - loss: 4.8489 - acc: 0.7000 - ETA: 2s - loss: 6.6621 - acc: 0.5750 - ETA: 2s - loss: 6.3497 - acc: 0.6000 - ETA: 2s - loss: 6.4258 - acc: 0.5955 - ETA: 2s - loss: 6.2923 - acc: 0.6052 - ETA: 2s - loss: 6.2106 - acc: 0.6111 - ETA: 2s - loss: 6.0805 - acc: 0.6198 - ETA: 2s - loss: 6.0065 - acc: 0.6240 - ETA: 2s - loss: 5.9357 - acc: 0.6281 - ETA: 2s - loss: 5.8964 - acc: 0.6289 - ETA: 2s - loss: 5.8611 - acc: 0.6317 - ETA: 1s - loss: 5.8011 - acc: 0.6348 - ETA: 1s - loss: 5.6953 - acc: 0.6413 - ETA: 1s - loss: 5.6485 - acc: 0.6446 - ETA: 1s - loss: 5.6341 - acc: 0.6455 - ETA: 1s - loss: 5.6424 - acc: 0.6454 - ETA: 1s - loss: 5.6654 - acc: 0.6435 - ETA: 1s - loss: 5.6789 - acc: 0.6427 - ETA: 1s - loss: 5.6905 - acc: 0.6416 - ETA: 1s - loss: 5.6190 - acc: 0.6457 - ETA: 1s - loss: 5.6331 - acc: 0.6445 - ETA: 1s - loss: 5.6456 - acc: 0.6435 - ETA: 1s - loss: 5.6631 - acc: 0.6425 - ETA: 1s - loss: 5.6240 - acc: 0.6452 - ETA: 1s - loss: 5.6268 - acc: 0.6440 - ETA: 1s - loss: 5.6594 - acc: 0.6418 - ETA: 1s - loss: 5.6994 - acc: 0.6392 - ETA: 0s - loss: 5.7048 - acc: 0.6386 - ETA: 0s - loss: 5.7074 - acc: 0.6383 - ETA: 0s - loss: 5.7019 - acc: 0.6388 - ETA: 0s - loss: 5.6966 - acc: 0.6394 - ETA: 0s - loss: 5.7306 - acc: 0.6373 - ETA: 0s - loss: 5.7256 - acc: 0.6373 - ETA: 0s - loss: 5.7371 - acc: 0.6364 - ETA: 0s - loss: 5.7020 - acc: 0.6388 - ETA: 0s - loss: 5.7346 - acc: 0.6368 - ETA: 0s - loss: 5.7413 - acc: 0.6366 - ETA: 0s - loss: 5.7549 - acc: 0.6358 - ETA: 0s - loss: 5.7557 - acc: 0.6357 - ETA: 0s - loss: 5.7392 - acc: 0.6367 - ETA: 0s - loss: 5.7404 - acc: 0.6363 - ETA: 0s - loss: 5.7598 - acc: 0.6353 - ETA: 0s - loss: 5.7652 - acc: 0.6350 - ETA: 0s - loss: 5.7667 - acc: 0.6347 - ETA: 0s - loss: 5.7741 - acc: 0.6344 - ETA: 0s - loss: 5.7790 - acc: 0.6341Epoch 00017: val_loss improved from 6.73259 to 6.72861, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 5.8033 - acc: 0.6326 - val_loss: 6.7286 - val_acc: 0.5102
    Epoch 19/20
    6580/6680 [============================>.] - ETA: 2s - loss: 7.2536 - acc: 0.5500 - ETA: 2s - loss: 6.2526 - acc: 0.6062 - ETA: 2s - loss: 6.5054 - acc: 0.5933 - ETA: 2s - loss: 6.3045 - acc: 0.6068 - ETA: 2s - loss: 6.0057 - acc: 0.6259 - ETA: 2s - loss: 6.0114 - acc: 0.6250 - ETA: 2s - loss: 6.0084 - acc: 0.6256 - ETA: 2s - loss: 5.7993 - acc: 0.6388 - ETA: 2s - loss: 5.7879 - acc: 0.6384 - ETA: 2s - loss: 5.8245 - acc: 0.6357 - ETA: 2s - loss: 5.8442 - acc: 0.6343 - ETA: 2s - loss: 5.8279 - acc: 0.6344 - ETA: 1s - loss: 5.8812 - acc: 0.6304 - ETA: 1s - loss: 5.9000 - acc: 0.6291 - ETA: 1s - loss: 5.8489 - acc: 0.6327 - ETA: 1s - loss: 5.8829 - acc: 0.6305 - ETA: 1s - loss: 5.9047 - acc: 0.6290 - ETA: 1s - loss: 5.8625 - acc: 0.6315 - ETA: 1s - loss: 5.8376 - acc: 0.6333 - ETA: 1s - loss: 5.8218 - acc: 0.6342 - ETA: 1s - loss: 5.8076 - acc: 0.6350 - ETA: 1s - loss: 5.8054 - acc: 0.6354 - ETA: 1s - loss: 5.8242 - acc: 0.6341 - ETA: 1s - loss: 5.8863 - acc: 0.6304 - ETA: 1s - loss: 5.8525 - acc: 0.6324 - ETA: 1s - loss: 5.8167 - acc: 0.6349 - ETA: 1s - loss: 5.7801 - acc: 0.6368 - ETA: 1s - loss: 5.7284 - acc: 0.6399 - ETA: 1s - loss: 5.7587 - acc: 0.6380 - ETA: 1s - loss: 5.7472 - acc: 0.6387 - ETA: 0s - loss: 5.7514 - acc: 0.6386 - ETA: 0s - loss: 5.7516 - acc: 0.6387 - ETA: 0s - loss: 5.7424 - acc: 0.6393 - ETA: 0s - loss: 5.7432 - acc: 0.6392 - ETA: 0s - loss: 5.7343 - acc: 0.6395 - ETA: 0s - loss: 5.7417 - acc: 0.6392 - ETA: 0s - loss: 5.7389 - acc: 0.6395 - ETA: 0s - loss: 5.7510 - acc: 0.6384 - ETA: 0s - loss: 5.7820 - acc: 0.6366 - ETA: 0s - loss: 5.7885 - acc: 0.6362 - ETA: 0s - loss: 5.7851 - acc: 0.6364 - ETA: 0s - loss: 5.7789 - acc: 0.6369 - ETA: 0s - loss: 5.7812 - acc: 0.6369 - ETA: 0s - loss: 5.7818 - acc: 0.6363 - ETA: 0s - loss: 5.7818 - acc: 0.6362 - ETA: 0s - loss: 5.7741 - acc: 0.6366 - ETA: 0s - loss: 5.7663 - acc: 0.6370 - ETA: 0s - loss: 5.7951 - acc: 0.6353Epoch 00018: val_loss did not improve
    6680/6680 [==============================] - 2s - loss: 5.7952 - acc: 0.6353 - val_loss: 6.7359 - val_acc: 0.5078
    Epoch 20/20
    6660/6680 [============================>.] - ETA: 2s - loss: 6.4483 - acc: 0.6000 - ETA: 2s - loss: 5.8747 - acc: 0.6312 - ETA: 2s - loss: 5.6598 - acc: 0.6467 - ETA: 2s - loss: 5.8374 - acc: 0.6364 - ETA: 2s - loss: 5.9014 - acc: 0.6328 - ETA: 2s - loss: 5.9978 - acc: 0.6270 - ETA: 2s - loss: 5.8129 - acc: 0.6386 - ETA: 2s - loss: 5.7894 - acc: 0.6402 - ETA: 1s - loss: 5.6879 - acc: 0.6466 - ETA: 1s - loss: 5.6588 - acc: 0.6485 - ETA: 1s - loss: 5.6683 - acc: 0.6480 - ETA: 1s - loss: 5.7082 - acc: 0.6451 - ETA: 1s - loss: 5.7317 - acc: 0.6432 - ETA: 1s - loss: 5.7760 - acc: 0.6405 - ETA: 1s - loss: 5.7825 - acc: 0.6398 - ETA: 1s - loss: 5.7809 - acc: 0.6400 - ETA: 1s - loss: 5.7718 - acc: 0.6407 - ETA: 1s - loss: 5.8484 - acc: 0.6360 - ETA: 1s - loss: 5.8743 - acc: 0.6345 - ETA: 1s - loss: 5.8974 - acc: 0.6331 - ETA: 1s - loss: 5.9129 - acc: 0.6322 - ETA: 1s - loss: 5.8964 - acc: 0.6330 - ETA: 1s - loss: 5.9509 - acc: 0.6297 - ETA: 1s - loss: 5.9387 - acc: 0.6299 - ETA: 1s - loss: 5.9285 - acc: 0.6306 - ETA: 1s - loss: 5.9429 - acc: 0.6292 - ETA: 1s - loss: 5.9608 - acc: 0.6276 - ETA: 0s - loss: 5.9545 - acc: 0.6279 - ETA: 0s - loss: 5.9095 - acc: 0.6308 - ETA: 0s - loss: 5.9242 - acc: 0.6298 - ETA: 0s - loss: 5.8855 - acc: 0.6323 - ETA: 0s - loss: 5.8491 - acc: 0.6346 - ETA: 0s - loss: 5.8292 - acc: 0.6357 - ETA: 0s - loss: 5.8127 - acc: 0.6363 - ETA: 0s - loss: 5.7917 - acc: 0.6378 - ETA: 0s - loss: 5.8227 - acc: 0.6359 - ETA: 0s - loss: 5.8534 - acc: 0.6339 - ETA: 0s - loss: 5.7925 - acc: 0.6374 - ETA: 0s - loss: 5.7795 - acc: 0.6378 - ETA: 0s - loss: 5.7950 - acc: 0.6368 - ETA: 0s - loss: 5.7993 - acc: 0.6363 - ETA: 0s - loss: 5.8122 - acc: 0.6356 - ETA: 0s - loss: 5.7920 - acc: 0.6366 - ETA: 0s - loss: 5.7849 - acc: 0.6369 - ETA: 0s - loss: 5.7492 - acc: 0.6389 - ETA: 0s - loss: 5.7653 - acc: 0.6378 - ETA: 0s - loss: 5.7805 - acc: 0.6367 - ETA: 0s - loss: 5.7801 - acc: 0.6368Epoch 00019: val_loss improved from 6.72861 to 6.69299, saving model to weights.best.VGG19.hdf5
    6680/6680 [==============================] - 2s - loss: 5.7797 - acc: 0.6368 - val_loss: 6.6930 - val_acc: 0.5030
    ---I am done saving model VGG19----
    
    In [39]:
    ### TODO: Train the model.
    checkpointer_Resnet50 = ModelCheckpoint(filepath='weights.best.Resnet50.hdf5', 
                                   verbose=1, save_best_only=True)
    
    Resnet50_model.fit(train_Resnet50, train_targets, 
              validation_data=(valid_Resnet50, valid_targets),
              epochs=20, batch_size=20, callbacks=[checkpointer_Resnet50], verbose=1)
    
    print('---I am done saving model valid_Resnet50 ----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/20
    6660/6680 [============================>.] - ETA: 444s - loss: 5.5372 - acc: 0.0000e+00 - ETA: 254s - loss: 6.2062 - acc: 0.0000e+00 - ETA: 192s - loss: 6.1238 - acc: 0.0000e+00 - ETA: 156s - loss: 6.1489 - acc: 0.0000e+00 - ETA: 136s - loss: 6.0100 - acc: 0.0100     - ETA: 122s - loss: 6.1365 - acc: 0.0167 - ETA: 113s - loss: 5.9950 - acc: 0.0214 - ETA: 105s - loss: 5.9432 - acc: 0.0250 - ETA: 97s - loss: 5.8589 - acc: 0.0278  - ETA: 93s - loss: 5.8087 - acc: 0.0350 - ETA: 88s - loss: 5.7267 - acc: 0.0409 - ETA: 83s - loss: 5.6160 - acc: 0.0458 - ETA: 79s - loss: 5.5145 - acc: 0.0538 - ETA: 76s - loss: 5.4653 - acc: 0.0607 - ETA: 74s - loss: 5.4222 - acc: 0.0633 - ETA: 70s - loss: 5.3523 - acc: 0.0719 - ETA: 69s - loss: 5.3008 - acc: 0.0735 - ETA: 67s - loss: 5.2598 - acc: 0.0750 - ETA: 65s - loss: 5.1933 - acc: 0.0789 - ETA: 63s - loss: 5.1840 - acc: 0.0850 - ETA: 62s - loss: 5.1180 - acc: 0.0857 - ETA: 61s - loss: 5.0712 - acc: 0.0909 - ETA: 60s - loss: 5.0306 - acc: 0.0935 - ETA: 59s - loss: 4.9681 - acc: 0.0979 - ETA: 57s - loss: 4.9047 - acc: 0.1020 - ETA: 56s - loss: 4.8512 - acc: 0.1096 - ETA: 55s - loss: 4.7863 - acc: 0.1222 - ETA: 55s - loss: 4.7349 - acc: 0.1196 - ETA: 54s - loss: 4.6977 - acc: 0.1207 - ETA: 53s - loss: 4.6440 - acc: 0.1317 - ETA: 52s - loss: 4.5972 - acc: 0.1355 - ETA: 51s - loss: 4.5620 - acc: 0.1391 - ETA: 50s - loss: 4.5083 - acc: 0.1424 - ETA: 49s - loss: 4.4786 - acc: 0.1471 - ETA: 49s - loss: 4.4803 - acc: 0.1457 - ETA: 48s - loss: 4.4502 - acc: 0.1542 - ETA: 47s - loss: 4.4161 - acc: 0.1554 - ETA: 46s - loss: 4.3737 - acc: 0.1618 - ETA: 46s - loss: 4.3196 - acc: 0.1744 - ETA: 45s - loss: 4.2720 - acc: 0.1813 - ETA: 45s - loss: 4.2353 - acc: 0.1817 - ETA: 44s - loss: 4.2123 - acc: 0.1869 - ETA: 44s - loss: 4.1692 - acc: 0.1942 - ETA: 43s - loss: 4.1334 - acc: 0.1977 - ETA: 43s - loss: 4.1067 - acc: 0.2011 - ETA: 42s - loss: 4.0839 - acc: 0.2033 - ETA: 42s - loss: 4.0452 - acc: 0.2074 - ETA: 41s - loss: 4.0084 - acc: 0.2104 - ETA: 41s - loss: 3.9841 - acc: 0.2153 - ETA: 40s - loss: 3.9469 - acc: 0.2200 - ETA: 40s - loss: 3.9176 - acc: 0.2216 - ETA: 40s - loss: 3.8936 - acc: 0.2240 - ETA: 39s - loss: 3.8637 - acc: 0.2311 - ETA: 39s - loss: 3.8415 - acc: 0.2343 - ETA: 39s - loss: 3.8185 - acc: 0.2382 - ETA: 39s - loss: 3.7887 - acc: 0.2438 - ETA: 38s - loss: 3.7632 - acc: 0.2491 - ETA: 38s - loss: 3.7495 - acc: 0.2509 - ETA: 37s - loss: 3.7144 - acc: 0.2568 - ETA: 37s - loss: 3.6813 - acc: 0.2617 - ETA: 37s - loss: 3.6597 - acc: 0.2656 - ETA: 37s - loss: 3.6488 - acc: 0.2653 - ETA: 36s - loss: 3.6274 - acc: 0.2675 - ETA: 36s - loss: 3.6029 - acc: 0.2711 - ETA: 36s - loss: 3.5972 - acc: 0.2700 - ETA: 35s - loss: 3.5764 - acc: 0.2727 - ETA: 35s - loss: 3.5558 - acc: 0.2746 - ETA: 35s - loss: 3.5297 - acc: 0.2816 - ETA: 35s - loss: 3.5144 - acc: 0.2855 - ETA: 34s - loss: 3.4845 - acc: 0.2907 - ETA: 34s - loss: 3.4361 - acc: 0.3000 - ETA: 33s - loss: 3.4142 - acc: 0.3021 - ETA: 33s - loss: 3.3942 - acc: 0.3047 - ETA: 33s - loss: 3.3748 - acc: 0.3073 - ETA: 33s - loss: 3.3572 - acc: 0.3092 - ETA: 32s - loss: 3.3398 - acc: 0.3117 - ETA: 32s - loss: 3.3224 - acc: 0.3135 - ETA: 32s - loss: 3.3032 - acc: 0.3171 - ETA: 31s - loss: 3.2849 - acc: 0.3206 - ETA: 31s - loss: 3.2686 - acc: 0.3228 - ETA: 31s - loss: 3.2560 - acc: 0.3238 - ETA: 31s - loss: 3.2397 - acc: 0.3277 - ETA: 30s - loss: 3.2255 - acc: 0.3292 - ETA: 30s - loss: 3.2121 - acc: 0.3312 - ETA: 30s - loss: 3.2049 - acc: 0.3314 - ETA: 30s - loss: 3.1815 - acc: 0.3368 - ETA: 29s - loss: 3.1565 - acc: 0.3426 - ETA: 29s - loss: 3.1386 - acc: 0.3461 - ETA: 29s - loss: 3.1284 - acc: 0.3456 - ETA: 29s - loss: 3.1180 - acc: 0.3456 - ETA: 28s - loss: 3.0993 - acc: 0.3484 - ETA: 28s - loss: 3.0870 - acc: 0.3500 - ETA: 28s - loss: 3.0746 - acc: 0.3521 - ETA: 28s - loss: 3.0623 - acc: 0.3532 - ETA: 27s - loss: 3.0474 - acc: 0.3542 - ETA: 27s - loss: 3.0377 - acc: 0.3552 - ETA: 27s - loss: 3.0227 - acc: 0.3566 - ETA: 27s - loss: 2.9854 - acc: 0.3640 - ETA: 26s - loss: 2.9736 - acc: 0.3653 - ETA: 26s - loss: 2.9587 - acc: 0.3686 - ETA: 26s - loss: 2.9427 - acc: 0.3723 - ETA: 26s - loss: 2.9255 - acc: 0.3760 - ETA: 26s - loss: 2.9117 - acc: 0.3767 - ETA: 25s - loss: 2.8980 - acc: 0.3797 - ETA: 25s - loss: 2.8757 - acc: 0.3815 - ETA: 24s - loss: 2.8454 - acc: 0.3859 - ETA: 24s - loss: 2.8346 - acc: 0.3883 - ETA: 24s - loss: 2.8231 - acc: 0.3897 - ETA: 24s - loss: 2.8110 - acc: 0.3907 - ETA: 23s - loss: 2.7956 - acc: 0.3930 - ETA: 23s - loss: 2.7720 - acc: 0.3970 - ETA: 23s - loss: 2.7564 - acc: 0.4004 - ETA: 22s - loss: 2.7351 - acc: 0.4029 - ETA: 22s - loss: 2.7121 - acc: 0.4066 - ETA: 21s - loss: 2.6941 - acc: 0.4085 - ETA: 21s - loss: 2.6702 - acc: 0.4115 - ETA: 21s - loss: 2.6625 - acc: 0.4126 - ETA: 20s - loss: 2.6480 - acc: 0.4155 - ETA: 20s - loss: 2.6247 - acc: 0.4206 - ETA: 19s - loss: 2.6056 - acc: 0.4237 - ETA: 19s - loss: 2.5821 - acc: 0.4281 - ETA: 18s - loss: 2.5511 - acc: 0.4333 - ETA: 18s - loss: 2.5282 - acc: 0.4386 - ETA: 18s - loss: 2.5001 - acc: 0.4434 - ETA: 17s - loss: 2.4704 - acc: 0.4486 - ETA: 17s - loss: 2.4654 - acc: 0.4493 - ETA: 16s - loss: 2.4370 - acc: 0.4540 - ETA: 16s - loss: 2.4131 - acc: 0.4578 - ETA: 15s - loss: 2.4020 - acc: 0.4594 - ETA: 15s - loss: 2.3843 - acc: 0.4618 - ETA: 15s - loss: 2.3649 - acc: 0.4657 - ETA: 14s - loss: 2.3430 - acc: 0.4691 - ETA: 14s - loss: 2.3247 - acc: 0.4735 - ETA: 13s - loss: 2.2990 - acc: 0.4774 - ETA: 13s - loss: 2.2754 - acc: 0.4808 - ETA: 12s - loss: 2.2588 - acc: 0.4829 - ETA: 12s - loss: 2.2415 - acc: 0.4854 - ETA: 12s - loss: 2.2240 - acc: 0.4876 - ETA: 11s - loss: 2.2064 - acc: 0.4902 - ETA: 11s - loss: 2.1941 - acc: 0.4928 - ETA: 10s - loss: 2.1736 - acc: 0.4963 - ETA: 10s - loss: 2.1545 - acc: 0.5013 - ETA: 10s - loss: 2.1346 - acc: 0.5053 - ETA: 9s - loss: 2.1157 - acc: 0.5095  - ETA: 9s - loss: 2.0986 - acc: 0.5118 - ETA: 9s - loss: 2.0825 - acc: 0.5148 - ETA: 8s - loss: 2.0668 - acc: 0.5172 - ETA: 8s - loss: 2.0470 - acc: 0.5204 - ETA: 8s - loss: 2.0365 - acc: 0.5221 - ETA: 8s - loss: 2.0217 - acc: 0.5245 - ETA: 7s - loss: 2.0047 - acc: 0.5266 - ETA: 7s - loss: 1.9837 - acc: 0.5310 - ETA: 7s - loss: 1.9740 - acc: 0.5331 - ETA: 6s - loss: 1.9619 - acc: 0.5353 - ETA: 6s - loss: 1.9512 - acc: 0.5362 - ETA: 6s - loss: 1.9272 - acc: 0.5404 - ETA: 5s - loss: 1.9177 - acc: 0.5432 - ETA: 5s - loss: 1.9025 - acc: 0.5457 - ETA: 5s - loss: 1.8874 - acc: 0.5488 - ETA: 5s - loss: 1.8738 - acc: 0.5516 - ETA: 4s - loss: 1.8559 - acc: 0.5552 - ETA: 4s - loss: 1.8427 - acc: 0.5569 - ETA: 4s - loss: 1.8278 - acc: 0.5598 - ETA: 3s - loss: 1.8145 - acc: 0.5622 - ETA: 3s - loss: 1.7965 - acc: 0.5667 - ETA: 3s - loss: 1.7857 - acc: 0.5689 - ETA: 3s - loss: 1.7744 - acc: 0.5710 - ETA: 2s - loss: 1.7567 - acc: 0.5748 - ETA: 2s - loss: 1.7429 - acc: 0.5771 - ETA: 2s - loss: 1.7244 - acc: 0.5808 - ETA: 1s - loss: 1.7088 - acc: 0.5844 - ETA: 1s - loss: 1.6991 - acc: 0.5862 - ETA: 1s - loss: 1.6830 - acc: 0.5891 - ETA: 0s - loss: 1.6670 - acc: 0.5920 - ETA: 0s - loss: 1.6561 - acc: 0.5939 - ETA: 0s - loss: 1.6458 - acc: 0.5953 - ETA: 0s - loss: 1.6297 - acc: 0.5988Epoch 00000: val_loss improved from inf to 0.79178, saving model to weights.best.Resnet50.hdf5
    6680/6680 [==============================] - 16s - loss: 1.6266 - acc: 0.5996 - val_loss: 0.7918 - val_acc: 0.7689
    Epoch 2/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.4916 - acc: 0.8500 - ETA: 3s - loss: 0.5504 - acc: 0.8143 - ETA: 3s - loss: 0.4600 - acc: 0.8542 - ETA: 3s - loss: 0.4171 - acc: 0.8735 - ETA: 3s - loss: 0.4052 - acc: 0.8773 - ETA: 3s - loss: 0.3841 - acc: 0.8852 - ETA: 3s - loss: 0.3836 - acc: 0.8844 - ETA: 3s - loss: 0.3810 - acc: 0.8811 - ETA: 3s - loss: 0.3809 - acc: 0.8798 - ETA: 2s - loss: 0.3859 - acc: 0.8777 - ETA: 2s - loss: 0.3895 - acc: 0.8750 - ETA: 2s - loss: 0.3994 - acc: 0.8737 - ETA: 2s - loss: 0.4030 - acc: 0.8742 - ETA: 2s - loss: 0.4064 - acc: 0.8701 - ETA: 2s - loss: 0.4108 - acc: 0.8688 - ETA: 2s - loss: 0.4144 - acc: 0.8669 - ETA: 2s - loss: 0.4131 - acc: 0.8663 - ETA: 2s - loss: 0.4122 - acc: 0.8657 - ETA: 2s - loss: 0.4259 - acc: 0.8617 - ETA: 2s - loss: 0.4322 - acc: 0.8606 - ETA: 2s - loss: 0.4270 - acc: 0.8639 - ETA: 2s - loss: 0.4329 - acc: 0.8628 - ETA: 2s - loss: 0.4290 - acc: 0.8645 - ETA: 2s - loss: 0.4341 - acc: 0.8634 - ETA: 2s - loss: 0.4328 - acc: 0.8637 - ETA: 2s - loss: 0.4330 - acc: 0.8640 - ETA: 2s - loss: 0.4293 - acc: 0.8649 - ETA: 1s - loss: 0.4306 - acc: 0.8644 - ETA: 1s - loss: 0.4316 - acc: 0.8656 - ETA: 1s - loss: 0.4307 - acc: 0.8651 - ETA: 1s - loss: 0.4284 - acc: 0.8653 - ETA: 1s - loss: 0.4288 - acc: 0.8654 - ETA: 1s - loss: 0.4276 - acc: 0.8662 - ETA: 1s - loss: 0.4296 - acc: 0.8663 - ETA: 1s - loss: 0.4293 - acc: 0.8660 - ETA: 1s - loss: 0.4301 - acc: 0.8652 - ETA: 1s - loss: 0.4319 - acc: 0.8647 - ETA: 1s - loss: 0.4326 - acc: 0.8646 - ETA: 1s - loss: 0.4364 - acc: 0.8632 - ETA: 1s - loss: 0.4382 - acc: 0.8631 - ETA: 1s - loss: 0.4415 - acc: 0.8623 - ETA: 1s - loss: 0.4413 - acc: 0.8627 - ETA: 1s - loss: 0.4433 - acc: 0.8622 - ETA: 1s - loss: 0.4413 - acc: 0.8635 - ETA: 1s - loss: 0.4402 - acc: 0.8639 - ETA: 1s - loss: 0.4399 - acc: 0.8640 - ETA: 0s - loss: 0.4439 - acc: 0.8637 - ETA: 0s - loss: 0.4437 - acc: 0.8636 - ETA: 0s - loss: 0.4462 - acc: 0.8623 - ETA: 0s - loss: 0.4500 - acc: 0.8615 - ETA: 0s - loss: 0.4531 - acc: 0.8607 - ETA: 0s - loss: 0.4553 - acc: 0.8605 - ETA: 0s - loss: 0.4556 - acc: 0.8609 - ETA: 0s - loss: 0.4557 - acc: 0.8606 - ETA: 0s - loss: 0.4534 - acc: 0.8609 - ETA: 0s - loss: 0.4533 - acc: 0.8611 - ETA: 0s - loss: 0.4543 - acc: 0.8602 - ETA: 0s - loss: 0.4521 - acc: 0.8605 - ETA: 0s - loss: 0.4510 - acc: 0.8610 - ETA: 0s - loss: 0.4493 - acc: 0.8618 - ETA: 0s - loss: 0.4508 - acc: 0.8615 - ETA: 0s - loss: 0.4496 - acc: 0.8622 - ETA: 0s - loss: 0.4513 - acc: 0.8611 - ETA: 0s - loss: 0.4503 - acc: 0.8618 - ETA: 0s - loss: 0.4490 - acc: 0.8618Epoch 00001: val_loss improved from 0.79178 to 0.71846, saving model to weights.best.Resnet50.hdf5
    6680/6680 [==============================] - 3s - loss: 0.4488 - acc: 0.8615 - val_loss: 0.7185 - val_acc: 0.7868
    Epoch 3/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.0401 - acc: 1.0000 - ETA: 3s - loss: 0.1818 - acc: 0.9583 - ETA: 3s - loss: 0.2350 - acc: 0.9273 - ETA: 3s - loss: 0.2162 - acc: 0.9375 - ETA: 3s - loss: 0.2038 - acc: 0.9409 - ETA: 3s - loss: 0.1937 - acc: 0.9464 - ETA: 2s - loss: 0.2025 - acc: 0.9441 - ETA: 2s - loss: 0.2118 - acc: 0.9385 - ETA: 2s - loss: 0.2247 - acc: 0.9341 - ETA: 2s - loss: 0.2152 - acc: 0.9388 - ETA: 2s - loss: 0.2203 - acc: 0.9333 - ETA: 2s - loss: 0.2185 - acc: 0.9331 - ETA: 2s - loss: 0.2272 - acc: 0.9305 - ETA: 2s - loss: 0.2390 - acc: 0.9261 - ETA: 2s - loss: 0.2410 - acc: 0.9243 - ETA: 2s - loss: 0.2481 - acc: 0.9215 - ETA: 2s - loss: 0.2507 - acc: 0.9214 - ETA: 2s - loss: 0.2508 - acc: 0.9213 - ETA: 2s - loss: 0.2504 - acc: 0.9218 - ETA: 2s - loss: 0.2449 - acc: 0.9247 - ETA: 2s - loss: 0.2508 - acc: 0.9221 - ETA: 2s - loss: 0.2520 - acc: 0.9216 - ETA: 2s - loss: 0.2496 - acc: 0.9224 - ETA: 2s - loss: 0.2486 - acc: 0.9227 - ETA: 2s - loss: 0.2456 - acc: 0.9238 - ETA: 2s - loss: 0.2430 - acc: 0.9236 - ETA: 2s - loss: 0.2419 - acc: 0.9237 - ETA: 1s - loss: 0.2368 - acc: 0.9252 - ETA: 1s - loss: 0.2348 - acc: 0.9255 - ETA: 1s - loss: 0.2367 - acc: 0.9245 - ETA: 1s - loss: 0.2374 - acc: 0.9239 - ETA: 1s - loss: 0.2421 - acc: 0.9226 - ETA: 1s - loss: 0.2418 - acc: 0.9231 - ETA: 1s - loss: 0.2408 - acc: 0.9230 - ETA: 1s - loss: 0.2413 - acc: 0.9223 - ETA: 1s - loss: 0.2437 - acc: 0.9220 - ETA: 1s - loss: 0.2424 - acc: 0.9228 - ETA: 1s - loss: 0.2438 - acc: 0.9228 - ETA: 1s - loss: 0.2426 - acc: 0.9230 - ETA: 1s - loss: 0.2446 - acc: 0.9232 - ETA: 1s - loss: 0.2431 - acc: 0.9236 - ETA: 1s - loss: 0.2470 - acc: 0.9221 - ETA: 1s - loss: 0.2503 - acc: 0.9209 - ETA: 1s - loss: 0.2519 - acc: 0.9202 - ETA: 1s - loss: 0.2549 - acc: 0.9189 - ETA: 0s - loss: 0.2544 - acc: 0.9189 - ETA: 0s - loss: 0.2540 - acc: 0.9185 - ETA: 0s - loss: 0.2550 - acc: 0.9176 - ETA: 0s - loss: 0.2561 - acc: 0.9176 - ETA: 0s - loss: 0.2568 - acc: 0.9175 - ETA: 0s - loss: 0.2568 - acc: 0.9177 - ETA: 0s - loss: 0.2585 - acc: 0.9177 - ETA: 0s - loss: 0.2607 - acc: 0.9178 - ETA: 0s - loss: 0.2633 - acc: 0.9169 - ETA: 0s - loss: 0.2639 - acc: 0.9164 - ETA: 0s - loss: 0.2642 - acc: 0.9164 - ETA: 0s - loss: 0.2639 - acc: 0.9163 - ETA: 0s - loss: 0.2646 - acc: 0.9163 - ETA: 0s - loss: 0.2643 - acc: 0.9164 - ETA: 0s - loss: 0.2654 - acc: 0.9163 - ETA: 0s - loss: 0.2653 - acc: 0.9162 - ETA: 0s - loss: 0.2663 - acc: 0.9158 - ETA: 0s - loss: 0.2646 - acc: 0.9162 - ETA: 0s - loss: 0.2647 - acc: 0.9162Epoch 00002: val_loss improved from 0.71846 to 0.63942, saving model to weights.best.Resnet50.hdf5
    6680/6680 [==============================] - 3s - loss: 0.2650 - acc: 0.9160 - val_loss: 0.6394 - val_acc: 0.8168
    Epoch 4/20
    6660/6680 [============================>.] - ETA: 3s - loss: 0.0682 - acc: 1.0000 - ETA: 3s - loss: 0.1643 - acc: 0.9500 - ETA: 3s - loss: 0.1523 - acc: 0.9500 - ETA: 3s - loss: 0.1714 - acc: 0.9556 - ETA: 3s - loss: 0.1455 - acc: 0.9630 - ETA: 3s - loss: 0.1397 - acc: 0.9655 - ETA: 2s - loss: 0.1493 - acc: 0.9603 - ETA: 2s - loss: 0.1496 - acc: 0.9603 - ETA: 2s - loss: 0.1406 - acc: 0.9636 - ETA: 2s - loss: 0.1550 - acc: 0.9541 - ETA: 2s - loss: 0.1580 - acc: 0.9537 - ETA: 2s - loss: 0.1625 - acc: 0.9542 - ETA: 2s - loss: 0.1609 - acc: 0.9545 - ETA: 2s - loss: 0.1636 - acc: 0.9535 - ETA: 2s - loss: 0.1653 - acc: 0.9500 - ETA: 2s - loss: 0.1648 - acc: 0.9500 - ETA: 2s - loss: 0.1646 - acc: 0.9506 - ETA: 2s - loss: 0.1654 - acc: 0.9495 - ETA: 2s - loss: 0.1673 - acc: 0.9490 - ETA: 2s - loss: 0.1690 - acc: 0.9481 - ETA: 2s - loss: 0.1650 - acc: 0.9491 - ETA: 2s - loss: 0.1646 - acc: 0.9491 - ETA: 2s - loss: 0.1615 - acc: 0.9504 - ETA: 2s - loss: 0.1621 - acc: 0.9512 - ETA: 1s - loss: 0.1617 - acc: 0.9504 - ETA: 1s - loss: 0.1593 - acc: 0.9519 - ETA: 1s - loss: 0.1634 - acc: 0.9504 - ETA: 1s - loss: 0.1635 - acc: 0.9503 - ETA: 1s - loss: 0.1641 - acc: 0.9507 - ETA: 1s - loss: 0.1630 - acc: 0.9513 - ETA: 1s - loss: 0.1626 - acc: 0.9506 - ETA: 1s - loss: 0.1653 - acc: 0.9500 - ETA: 1s - loss: 0.1665 - acc: 0.9497 - ETA: 1s - loss: 0.1676 - acc: 0.9489 - ETA: 1s - loss: 0.1655 - acc: 0.9500 - ETA: 1s - loss: 0.1668 - acc: 0.9497 - ETA: 1s - loss: 0.1691 - acc: 0.9489 - ETA: 1s - loss: 0.1684 - acc: 0.9495 - ETA: 1s - loss: 0.1668 - acc: 0.9500 - ETA: 1s - loss: 0.1651 - acc: 0.9505 - ETA: 1s - loss: 0.1633 - acc: 0.9510 - ETA: 1s - loss: 0.1614 - acc: 0.9512 - ETA: 1s - loss: 0.1607 - acc: 0.9511 - ETA: 1s - loss: 0.1610 - acc: 0.9511 - ETA: 1s - loss: 0.1623 - acc: 0.9509 - ETA: 0s - loss: 0.1637 - acc: 0.9502 - ETA: 0s - loss: 0.1616 - acc: 0.9506 - ETA: 0s - loss: 0.1637 - acc: 0.9498 - ETA: 0s - loss: 0.1643 - acc: 0.9498 - ETA: 0s - loss: 0.1638 - acc: 0.9500 - ETA: 0s - loss: 0.1642 - acc: 0.9496 - ETA: 0s - loss: 0.1634 - acc: 0.9500 - ETA: 0s - loss: 0.1643 - acc: 0.9500 - ETA: 0s - loss: 0.1654 - acc: 0.9498 - ETA: 0s - loss: 0.1676 - acc: 0.9490 - ETA: 0s - loss: 0.1681 - acc: 0.9488 - ETA: 0s - loss: 0.1700 - acc: 0.9482 - ETA: 0s - loss: 0.1731 - acc: 0.9472 - ETA: 0s - loss: 0.1749 - acc: 0.9471 - ETA: 0s - loss: 0.1763 - acc: 0.9465 - ETA: 0s - loss: 0.1761 - acc: 0.9465 - ETA: 0s - loss: 0.1768 - acc: 0.9463 - ETA: 0s - loss: 0.1769 - acc: 0.9462 - ETA: 0s - loss: 0.1763 - acc: 0.9465Epoch 00003: val_loss improved from 0.63942 to 0.62551, saving model to weights.best.Resnet50.hdf5
    6680/6680 [==============================] - 3s - loss: 0.1768 - acc: 0.9466 - val_loss: 0.6255 - val_acc: 0.8192
    Epoch 5/20
    6660/6680 [============================>.] - ETA: 3s - loss: 0.1726 - acc: 1.0000 - ETA: 3s - loss: 0.0575 - acc: 1.0000 - ETA: 3s - loss: 0.0817 - acc: 0.9909 - ETA: 3s - loss: 0.0840 - acc: 0.9906 - ETA: 3s - loss: 0.0841 - acc: 0.9905 - ETA: 3s - loss: 0.0838 - acc: 0.9846 - ETA: 3s - loss: 0.0810 - acc: 0.9855 - ETA: 3s - loss: 0.0788 - acc: 0.9861 - ETA: 2s - loss: 0.0798 - acc: 0.9845 - ETA: 2s - loss: 0.0917 - acc: 0.9787 - ETA: 2s - loss: 0.0927 - acc: 0.9774 - ETA: 2s - loss: 0.0917 - acc: 0.9763 - ETA: 2s - loss: 0.0902 - acc: 0.9754 - ETA: 2s - loss: 0.0920 - acc: 0.9739 - ETA: 2s - loss: 0.0883 - acc: 0.9753 - ETA: 2s - loss: 0.0888 - acc: 0.9747 - ETA: 2s - loss: 0.0951 - acc: 0.9713 - ETA: 2s - loss: 0.0936 - acc: 0.9721 - ETA: 2s - loss: 0.0919 - acc: 0.9728 - ETA: 2s - loss: 0.0935 - acc: 0.9722 - ETA: 2s - loss: 0.0980 - acc: 0.9712 - ETA: 2s - loss: 0.0986 - acc: 0.9703 - ETA: 2s - loss: 0.0976 - acc: 0.9702 - ETA: 2s - loss: 0.0990 - acc: 0.9694 - ETA: 1s - loss: 0.1014 - acc: 0.9683 - ETA: 1s - loss: 0.1034 - acc: 0.9680 - ETA: 1s - loss: 0.1047 - acc: 0.9674 - ETA: 1s - loss: 0.1062 - acc: 0.9675 - ETA: 1s - loss: 0.1063 - acc: 0.9675 - ETA: 1s - loss: 0.1056 - acc: 0.9679 - ETA: 1s - loss: 0.1058 - acc: 0.9677 - ETA: 1s - loss: 0.1060 - acc: 0.9675 - ETA: 1s - loss: 0.1069 - acc: 0.9675 - ETA: 1s - loss: 0.1053 - acc: 0.9676 - ETA: 1s - loss: 0.1059 - acc: 0.9669 - ETA: 1s - loss: 0.1051 - acc: 0.9669 - ETA: 1s - loss: 0.1052 - acc: 0.9670 - ETA: 1s - loss: 0.1067 - acc: 0.9666 - ETA: 1s - loss: 0.1101 - acc: 0.9654 - ETA: 1s - loss: 0.1129 - acc: 0.9648 - ETA: 1s - loss: 0.1136 - acc: 0.9644 - ETA: 1s - loss: 0.1147 - acc: 0.9641 - ETA: 1s - loss: 0.1147 - acc: 0.9637 - ETA: 1s - loss: 0.1157 - acc: 0.9632 - ETA: 1s - loss: 0.1151 - acc: 0.9634 - ETA: 0s - loss: 0.1145 - acc: 0.9635 - ETA: 0s - loss: 0.1158 - acc: 0.9632 - ETA: 0s - loss: 0.1157 - acc: 0.9636 - ETA: 0s - loss: 0.1148 - acc: 0.9640 - ETA: 0s - loss: 0.1171 - acc: 0.9633 - ETA: 0s - loss: 0.1196 - acc: 0.9628 - ETA: 0s - loss: 0.1201 - acc: 0.9622 - ETA: 0s - loss: 0.1216 - acc: 0.9616 - ETA: 0s - loss: 0.1229 - acc: 0.9610 - ETA: 0s - loss: 0.1222 - acc: 0.9611 - ETA: 0s - loss: 0.1224 - acc: 0.9614 - ETA: 0s - loss: 0.1209 - acc: 0.9618 - ETA: 0s - loss: 0.1218 - acc: 0.9617 - ETA: 0s - loss: 0.1257 - acc: 0.9604 - ETA: 0s - loss: 0.1261 - acc: 0.9605 - ETA: 0s - loss: 0.1274 - acc: 0.9608 - ETA: 0s - loss: 0.1272 - acc: 0.9608 - ETA: 0s - loss: 0.1268 - acc: 0.9608Epoch 00004: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.1271 - acc: 0.9608 - val_loss: 0.7038 - val_acc: 0.8096
    Epoch 6/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.1406 - acc: 0.9500 - ETA: 3s - loss: 0.0521 - acc: 0.9833 - ETA: 3s - loss: 0.0800 - acc: 0.9773 - ETA: 3s - loss: 0.0676 - acc: 0.9781 - ETA: 3s - loss: 0.0613 - acc: 0.9810 - ETA: 3s - loss: 0.0593 - acc: 0.9827 - ETA: 3s - loss: 0.0598 - acc: 0.9806 - ETA: 3s - loss: 0.0694 - acc: 0.9792 - ETA: 2s - loss: 0.0666 - acc: 0.9793 - ETA: 2s - loss: 0.0659 - acc: 0.9793 - ETA: 2s - loss: 0.0704 - acc: 0.9784 - ETA: 2s - loss: 0.0666 - acc: 0.9804 - ETA: 2s - loss: 0.0695 - acc: 0.9795 - ETA: 2s - loss: 0.0686 - acc: 0.9803 - ETA: 2s - loss: 0.0711 - acc: 0.9796 - ETA: 2s - loss: 0.0685 - acc: 0.9809 - ETA: 2s - loss: 0.0702 - acc: 0.9796 - ETA: 2s - loss: 0.0687 - acc: 0.9802 - ETA: 2s - loss: 0.0731 - acc: 0.9786 - ETA: 2s - loss: 0.0763 - acc: 0.9781 - ETA: 2s - loss: 0.0751 - acc: 0.9782 - ETA: 2s - loss: 0.0751 - acc: 0.9778 - ETA: 2s - loss: 0.0740 - acc: 0.9779 - ETA: 2s - loss: 0.0731 - acc: 0.9776 - ETA: 2s - loss: 0.0738 - acc: 0.9773 - ETA: 2s - loss: 0.0735 - acc: 0.9778 - ETA: 2s - loss: 0.0731 - acc: 0.9784 - ETA: 1s - loss: 0.0737 - acc: 0.9786 - ETA: 1s - loss: 0.0726 - acc: 0.9792 - ETA: 1s - loss: 0.0709 - acc: 0.9797 - ETA: 1s - loss: 0.0706 - acc: 0.9798 - ETA: 1s - loss: 0.0704 - acc: 0.9796 - ETA: 1s - loss: 0.0709 - acc: 0.9795 - ETA: 1s - loss: 0.0712 - acc: 0.9796 - ETA: 1s - loss: 0.0702 - acc: 0.9803 - ETA: 1s - loss: 0.0717 - acc: 0.9798 - ETA: 1s - loss: 0.0739 - acc: 0.9786 - ETA: 1s - loss: 0.0765 - acc: 0.9783 - ETA: 1s - loss: 0.0777 - acc: 0.9777 - ETA: 1s - loss: 0.0783 - acc: 0.9774 - ETA: 1s - loss: 0.0793 - acc: 0.9770 - ETA: 1s - loss: 0.0807 - acc: 0.9761 - ETA: 1s - loss: 0.0818 - acc: 0.9760 - ETA: 1s - loss: 0.0824 - acc: 0.9759 - ETA: 0s - loss: 0.0817 - acc: 0.9762 - ETA: 0s - loss: 0.0808 - acc: 0.9765 - ETA: 0s - loss: 0.0801 - acc: 0.9767 - ETA: 0s - loss: 0.0807 - acc: 0.9764 - ETA: 0s - loss: 0.0805 - acc: 0.9763 - ETA: 0s - loss: 0.0804 - acc: 0.9765 - ETA: 0s - loss: 0.0802 - acc: 0.9766 - ETA: 0s - loss: 0.0807 - acc: 0.9761 - ETA: 0s - loss: 0.0811 - acc: 0.9760 - ETA: 0s - loss: 0.0829 - acc: 0.9757 - ETA: 0s - loss: 0.0830 - acc: 0.9758 - ETA: 0s - loss: 0.0828 - acc: 0.9757 - ETA: 0s - loss: 0.0827 - acc: 0.9758 - ETA: 0s - loss: 0.0829 - acc: 0.9755 - ETA: 0s - loss: 0.0838 - acc: 0.9749 - ETA: 0s - loss: 0.0856 - acc: 0.9747 - ETA: 0s - loss: 0.0856 - acc: 0.9746 - ETA: 0s - loss: 0.0854 - acc: 0.9745 - ETA: 0s - loss: 0.0866 - acc: 0.9743 - ETA: 0s - loss: 0.0862 - acc: 0.9745Epoch 00005: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0855 - acc: 0.9749 - val_loss: 0.7020 - val_acc: 0.8204
    Epoch 7/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.1021 - acc: 0.9000 - ETA: 3s - loss: 0.0550 - acc: 0.9750 - ETA: 3s - loss: 0.0449 - acc: 0.9818 - ETA: 3s - loss: 0.0382 - acc: 0.9875 - ETA: 3s - loss: 0.0447 - acc: 0.9881 - ETA: 3s - loss: 0.0463 - acc: 0.9846 - ETA: 3s - loss: 0.0442 - acc: 0.9855 - ETA: 3s - loss: 0.0507 - acc: 0.9833 - ETA: 3s - loss: 0.0495 - acc: 0.9829 - ETA: 2s - loss: 0.0449 - acc: 0.9851 - ETA: 2s - loss: 0.0430 - acc: 0.9856 - ETA: 2s - loss: 0.0419 - acc: 0.9862 - ETA: 2s - loss: 0.0451 - acc: 0.9852 - ETA: 2s - loss: 0.0464 - acc: 0.9850 - ETA: 2s - loss: 0.0473 - acc: 0.9849 - ETA: 2s - loss: 0.0476 - acc: 0.9854 - ETA: 2s - loss: 0.0473 - acc: 0.9852 - ETA: 2s - loss: 0.0462 - acc: 0.9856 - ETA: 2s - loss: 0.0482 - acc: 0.9845 - ETA: 2s - loss: 0.0486 - acc: 0.9844 - ETA: 2s - loss: 0.0470 - acc: 0.9853 - ETA: 2s - loss: 0.0484 - acc: 0.9847 - ETA: 2s - loss: 0.0485 - acc: 0.9847 - ETA: 1s - loss: 0.0473 - acc: 0.9854 - ETA: 1s - loss: 0.0471 - acc: 0.9860 - ETA: 1s - loss: 0.0494 - acc: 0.9849 - ETA: 1s - loss: 0.0500 - acc: 0.9845 - ETA: 1s - loss: 0.0489 - acc: 0.9851 - ETA: 1s - loss: 0.0511 - acc: 0.9840 - ETA: 1s - loss: 0.0506 - acc: 0.9841 - ETA: 1s - loss: 0.0510 - acc: 0.9837 - ETA: 1s - loss: 0.0505 - acc: 0.9839 - ETA: 1s - loss: 0.0506 - acc: 0.9838 - ETA: 1s - loss: 0.0527 - acc: 0.9840 - ETA: 1s - loss: 0.0550 - acc: 0.9832 - ETA: 1s - loss: 0.0553 - acc: 0.9828 - ETA: 1s - loss: 0.0561 - acc: 0.9825 - ETA: 1s - loss: 0.0557 - acc: 0.9824 - ETA: 1s - loss: 0.0555 - acc: 0.9826 - ETA: 1s - loss: 0.0560 - acc: 0.9826 - ETA: 1s - loss: 0.0592 - acc: 0.9816 - ETA: 1s - loss: 0.0591 - acc: 0.9813 - ETA: 1s - loss: 0.0589 - acc: 0.9815 - ETA: 0s - loss: 0.0588 - acc: 0.9815 - ETA: 0s - loss: 0.0579 - acc: 0.9819 - ETA: 0s - loss: 0.0576 - acc: 0.9818 - ETA: 0s - loss: 0.0582 - acc: 0.9816 - ETA: 0s - loss: 0.0593 - acc: 0.9810 - ETA: 0s - loss: 0.0594 - acc: 0.9810 - ETA: 0s - loss: 0.0591 - acc: 0.9809 - ETA: 0s - loss: 0.0596 - acc: 0.9809 - ETA: 0s - loss: 0.0614 - acc: 0.9802 - ETA: 0s - loss: 0.0608 - acc: 0.9805 - ETA: 0s - loss: 0.0614 - acc: 0.9805 - ETA: 0s - loss: 0.0619 - acc: 0.9807 - ETA: 0s - loss: 0.0639 - acc: 0.9797 - ETA: 0s - loss: 0.0645 - acc: 0.9792 - ETA: 0s - loss: 0.0639 - acc: 0.9793 - ETA: 0s - loss: 0.0642 - acc: 0.9794 - ETA: 0s - loss: 0.0639 - acc: 0.9794 - ETA: 0s - loss: 0.0633 - acc: 0.9797 - ETA: 0s - loss: 0.0634 - acc: 0.9797 - ETA: 0s - loss: 0.0635 - acc: 0.9795Epoch 00006: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0633 - acc: 0.9796 - val_loss: 0.7433 - val_acc: 0.8108
    Epoch 8/20
    6580/6680 [============================>.] - ETA: 18s - loss: 0.0726 - acc: 0.9500 - ETA: 5s - loss: 0.0217 - acc: 0.9929  - ETA: 3s - loss: 0.0331 - acc: 0.9923 - ETA: 3s - loss: 0.0492 - acc: 0.9868 - ETA: 3s - loss: 0.0451 - acc: 0.9880 - ETA: 3s - loss: 0.0449 - acc: 0.9871 - ETA: 3s - loss: 0.0446 - acc: 0.9865 - ETA: 2s - loss: 0.0407 - acc: 0.9884 - ETA: 2s - loss: 0.0414 - acc: 0.9878 - ETA: 2s - loss: 0.0384 - acc: 0.9882 - ETA: 2s - loss: 0.0375 - acc: 0.9877 - ETA: 2s - loss: 0.0367 - acc: 0.9888 - ETA: 2s - loss: 0.0376 - acc: 0.9877 - ETA: 2s - loss: 0.0382 - acc: 0.9880 - ETA: 2s - loss: 0.0392 - acc: 0.9876 - ETA: 2s - loss: 0.0408 - acc: 0.9863 - ETA: 2s - loss: 0.0392 - acc: 0.9871 - ETA: 2s - loss: 0.0376 - acc: 0.9879 - ETA: 2s - loss: 0.0426 - acc: 0.9876 - ETA: 2s - loss: 0.0418 - acc: 0.9882 - ETA: 2s - loss: 0.0410 - acc: 0.9887 - ETA: 1s - loss: 0.0398 - acc: 0.9891 - ETA: 1s - loss: 0.0414 - acc: 0.9880 - ETA: 1s - loss: 0.0407 - acc: 0.9884 - ETA: 1s - loss: 0.0414 - acc: 0.9881 - ETA: 1s - loss: 0.0408 - acc: 0.9885 - ETA: 1s - loss: 0.0407 - acc: 0.9883 - ETA: 1s - loss: 0.0407 - acc: 0.9883 - ETA: 1s - loss: 0.0412 - acc: 0.9881 - ETA: 1s - loss: 0.0417 - acc: 0.9881 - ETA: 1s - loss: 0.0413 - acc: 0.9885 - ETA: 1s - loss: 0.0405 - acc: 0.9888 - ETA: 1s - loss: 0.0404 - acc: 0.9888 - ETA: 1s - loss: 0.0397 - acc: 0.9889 - ETA: 1s - loss: 0.0398 - acc: 0.9886 - ETA: 1s - loss: 0.0403 - acc: 0.9884 - ETA: 1s - loss: 0.0397 - acc: 0.9887 - ETA: 1s - loss: 0.0393 - acc: 0.9890 - ETA: 1s - loss: 0.0392 - acc: 0.9890 - ETA: 1s - loss: 0.0391 - acc: 0.9893 - ETA: 1s - loss: 0.0389 - acc: 0.9893 - ETA: 1s - loss: 0.0401 - acc: 0.9891 - ETA: 1s - loss: 0.0405 - acc: 0.9886 - ETA: 0s - loss: 0.0404 - acc: 0.9887 - ETA: 0s - loss: 0.0407 - acc: 0.9885 - ETA: 0s - loss: 0.0404 - acc: 0.9887 - ETA: 0s - loss: 0.0417 - acc: 0.9886 - ETA: 0s - loss: 0.0416 - acc: 0.9886 - ETA: 0s - loss: 0.0424 - acc: 0.9880 - ETA: 0s - loss: 0.0423 - acc: 0.9881 - ETA: 0s - loss: 0.0443 - acc: 0.9874 - ETA: 0s - loss: 0.0450 - acc: 0.9869 - ETA: 0s - loss: 0.0447 - acc: 0.9869 - ETA: 0s - loss: 0.0443 - acc: 0.9871 - ETA: 0s - loss: 0.0448 - acc: 0.9870 - ETA: 0s - loss: 0.0448 - acc: 0.9869 - ETA: 0s - loss: 0.0450 - acc: 0.9870 - ETA: 0s - loss: 0.0460 - acc: 0.9868 - ETA: 0s - loss: 0.0462 - acc: 0.9869 - ETA: 0s - loss: 0.0463 - acc: 0.9869 - ETA: 0s - loss: 0.0461 - acc: 0.9870 - ETA: 0s - loss: 0.0465 - acc: 0.9869 - ETA: 0s - loss: 0.0464 - acc: 0.9869Epoch 00007: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0463 - acc: 0.9868 - val_loss: 0.7082 - val_acc: 0.8180
    Epoch 9/20
    6640/6680 [============================>.] - ETA: 3s - loss: 0.0081 - acc: 1.0000 - ETA: 2s - loss: 0.0227 - acc: 0.9929 - ETA: 2s - loss: 0.0189 - acc: 0.9962 - ETA: 2s - loss: 0.0186 - acc: 0.9974 - ETA: 2s - loss: 0.0191 - acc: 0.9980 - ETA: 2s - loss: 0.0191 - acc: 0.9984 - ETA: 2s - loss: 0.0199 - acc: 0.9973 - ETA: 2s - loss: 0.0261 - acc: 0.9953 - ETA: 2s - loss: 0.0256 - acc: 0.9959 - ETA: 2s - loss: 0.0275 - acc: 0.9936 - ETA: 2s - loss: 0.0276 - acc: 0.9934 - ETA: 2s - loss: 0.0283 - acc: 0.9925 - ETA: 2s - loss: 0.0285 - acc: 0.9925 - ETA: 2s - loss: 0.0340 - acc: 0.9905 - ETA: 2s - loss: 0.0343 - acc: 0.9905 - ETA: 2s - loss: 0.0341 - acc: 0.9900 - ETA: 2s - loss: 0.0332 - acc: 0.9906 - ETA: 2s - loss: 0.0342 - acc: 0.9907 - ETA: 2s - loss: 0.0349 - acc: 0.9903 - ETA: 1s - loss: 0.0336 - acc: 0.9908 - ETA: 1s - loss: 0.0330 - acc: 0.9908 - ETA: 1s - loss: 0.0321 - acc: 0.9913 - ETA: 1s - loss: 0.0329 - acc: 0.9913 - ETA: 1s - loss: 0.0343 - acc: 0.9909 - ETA: 1s - loss: 0.0341 - acc: 0.9909 - ETA: 1s - loss: 0.0339 - acc: 0.9909 - ETA: 1s - loss: 0.0339 - acc: 0.9908 - ETA: 1s - loss: 0.0339 - acc: 0.9908 - ETA: 1s - loss: 0.0331 - acc: 0.9911 - ETA: 1s - loss: 0.0322 - acc: 0.9914 - ETA: 1s - loss: 0.0316 - acc: 0.9917 - ETA: 1s - loss: 0.0328 - acc: 0.9911 - ETA: 1s - loss: 0.0326 - acc: 0.9910 - ETA: 1s - loss: 0.0324 - acc: 0.9910 - ETA: 1s - loss: 0.0321 - acc: 0.9910 - ETA: 1s - loss: 0.0317 - acc: 0.9912 - ETA: 1s - loss: 0.0316 - acc: 0.9912 - ETA: 1s - loss: 0.0331 - acc: 0.9912 - ETA: 1s - loss: 0.0331 - acc: 0.9914 - ETA: 1s - loss: 0.0327 - acc: 0.9916 - ETA: 1s - loss: 0.0328 - acc: 0.9913 - ETA: 0s - loss: 0.0328 - acc: 0.9913 - ETA: 0s - loss: 0.0325 - acc: 0.9914 - ETA: 0s - loss: 0.0321 - acc: 0.9915 - ETA: 0s - loss: 0.0322 - acc: 0.9915 - ETA: 0s - loss: 0.0317 - acc: 0.9917 - ETA: 0s - loss: 0.0314 - acc: 0.9918 - ETA: 0s - loss: 0.0313 - acc: 0.9918 - ETA: 0s - loss: 0.0312 - acc: 0.9918 - ETA: 0s - loss: 0.0313 - acc: 0.9917 - ETA: 0s - loss: 0.0317 - acc: 0.9915 - ETA: 0s - loss: 0.0313 - acc: 0.9917 - ETA: 0s - loss: 0.0312 - acc: 0.9916 - ETA: 0s - loss: 0.0316 - acc: 0.9914 - ETA: 0s - loss: 0.0328 - acc: 0.9911 - ETA: 0s - loss: 0.0327 - acc: 0.9912 - ETA: 0s - loss: 0.0327 - acc: 0.9912 - ETA: 0s - loss: 0.0338 - acc: 0.9907 - ETA: 0s - loss: 0.0346 - acc: 0.9904 - ETA: 0s - loss: 0.0344 - acc: 0.9904 - ETA: 0s - loss: 0.0347 - acc: 0.9901 - ETA: 0s - loss: 0.0351 - acc: 0.9899Epoch 00008: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0350 - acc: 0.9900 - val_loss: 0.7075 - val_acc: 0.8228
    Epoch 10/20
    6640/6680 [============================>.] - ETA: 3s - loss: 0.0046 - acc: 1.0000 - ETA: 3s - loss: 0.0273 - acc: 0.9929 - ETA: 3s - loss: 0.0175 - acc: 0.9958 - ETA: 3s - loss: 0.0348 - acc: 0.9912 - ETA: 3s - loss: 0.0307 - acc: 0.9909 - ETA: 3s - loss: 0.0267 - acc: 0.9926 - ETA: 3s - loss: 0.0305 - acc: 0.9906 - ETA: 3s - loss: 0.0271 - acc: 0.9919 - ETA: 2s - loss: 0.0279 - acc: 0.9917 - ETA: 2s - loss: 0.0275 - acc: 0.9927 - ETA: 2s - loss: 0.0255 - acc: 0.9935 - ETA: 2s - loss: 0.0236 - acc: 0.9942 - ETA: 2s - loss: 0.0220 - acc: 0.9947 - ETA: 2s - loss: 0.0216 - acc: 0.9951 - ETA: 2s - loss: 0.0211 - acc: 0.9955 - ETA: 2s - loss: 0.0205 - acc: 0.9958 - ETA: 2s - loss: 0.0198 - acc: 0.9961 - ETA: 2s - loss: 0.0202 - acc: 0.9958 - ETA: 2s - loss: 0.0198 - acc: 0.9961 - ETA: 2s - loss: 0.0199 - acc: 0.9958 - ETA: 2s - loss: 0.0199 - acc: 0.9952 - ETA: 1s - loss: 0.0198 - acc: 0.9954 - ETA: 1s - loss: 0.0199 - acc: 0.9952 - ETA: 1s - loss: 0.0210 - acc: 0.9951 - ETA: 1s - loss: 0.0214 - acc: 0.9949 - ETA: 1s - loss: 0.0217 - acc: 0.9948 - ETA: 1s - loss: 0.0225 - acc: 0.9943 - ETA: 1s - loss: 0.0225 - acc: 0.9942 - ETA: 1s - loss: 0.0241 - acc: 0.9941 - ETA: 1s - loss: 0.0246 - acc: 0.9940 - ETA: 1s - loss: 0.0243 - acc: 0.9943 - ETA: 1s - loss: 0.0241 - acc: 0.9942 - ETA: 1s - loss: 0.0253 - acc: 0.9938 - ETA: 1s - loss: 0.0256 - acc: 0.9935 - ETA: 1s - loss: 0.0255 - acc: 0.9934 - ETA: 1s - loss: 0.0249 - acc: 0.9936 - ETA: 1s - loss: 0.0246 - acc: 0.9938 - ETA: 1s - loss: 0.0256 - acc: 0.9933 - ETA: 1s - loss: 0.0263 - acc: 0.9928 - ETA: 0s - loss: 0.0271 - acc: 0.9925 - ETA: 0s - loss: 0.0270 - acc: 0.9925 - ETA: 0s - loss: 0.0269 - acc: 0.9925 - ETA: 0s - loss: 0.0266 - acc: 0.9925 - ETA: 0s - loss: 0.0267 - acc: 0.9924 - ETA: 0s - loss: 0.0267 - acc: 0.9924 - ETA: 0s - loss: 0.0266 - acc: 0.9925 - ETA: 0s - loss: 0.0267 - acc: 0.9927 - ETA: 0s - loss: 0.0274 - acc: 0.9924 - ETA: 0s - loss: 0.0270 - acc: 0.9926 - ETA: 0s - loss: 0.0268 - acc: 0.9927 - ETA: 0s - loss: 0.0272 - acc: 0.9927 - ETA: 0s - loss: 0.0276 - acc: 0.9926 - ETA: 0s - loss: 0.0275 - acc: 0.9927 - ETA: 0s - loss: 0.0272 - acc: 0.9929 - ETA: 0s - loss: 0.0269 - acc: 0.9930 - ETA: 0s - loss: 0.0267 - acc: 0.9931 - ETA: 0s - loss: 0.0266 - acc: 0.9932 - ETA: 0s - loss: 0.0269 - acc: 0.9932 - ETA: 0s - loss: 0.0267 - acc: 0.9933 - ETA: 0s - loss: 0.0267 - acc: 0.9931Epoch 00009: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0266 - acc: 0.9931 - val_loss: 0.7837 - val_acc: 0.8132
    Epoch 11/20
    6560/6680 [============================>.] - ETA: 3s - loss: 0.0026 - acc: 1.0000 - ETA: 3s - loss: 0.0070 - acc: 1.0000 - ETA: 3s - loss: 0.0065 - acc: 1.0000 - ETA: 3s - loss: 0.0125 - acc: 0.9971 - ETA: 3s - loss: 0.0125 - acc: 0.9955 - ETA: 3s - loss: 0.0122 - acc: 0.9944 - ETA: 3s - loss: 0.0127 - acc: 0.9937 - ETA: 2s - loss: 0.0122 - acc: 0.9946 - ETA: 2s - loss: 0.0130 - acc: 0.9940 - ETA: 2s - loss: 0.0146 - acc: 0.9936 - ETA: 2s - loss: 0.0203 - acc: 0.9933 - ETA: 2s - loss: 0.0193 - acc: 0.9939 - ETA: 2s - loss: 0.0182 - acc: 0.9944 - ETA: 2s - loss: 0.0175 - acc: 0.9948 - ETA: 2s - loss: 0.0192 - acc: 0.9944 - ETA: 2s - loss: 0.0183 - acc: 0.9948 - ETA: 2s - loss: 0.0177 - acc: 0.9951 - ETA: 2s - loss: 0.0173 - acc: 0.9954 - ETA: 2s - loss: 0.0175 - acc: 0.9957 - ETA: 2s - loss: 0.0176 - acc: 0.9954 - ETA: 2s - loss: 0.0188 - acc: 0.9946 - ETA: 2s - loss: 0.0186 - acc: 0.9949 - ETA: 2s - loss: 0.0181 - acc: 0.9951 - ETA: 2s - loss: 0.0184 - acc: 0.9953 - ETA: 2s - loss: 0.0181 - acc: 0.9955 - ETA: 2s - loss: 0.0183 - acc: 0.9953 - ETA: 2s - loss: 0.0182 - acc: 0.9955 - ETA: 2s - loss: 0.0180 - acc: 0.9956 - ETA: 1s - loss: 0.0177 - acc: 0.9958 - ETA: 1s - loss: 0.0178 - acc: 0.9956 - ETA: 1s - loss: 0.0181 - acc: 0.9954 - ETA: 1s - loss: 0.0180 - acc: 0.9955 - ETA: 1s - loss: 0.0191 - acc: 0.9954 - ETA: 1s - loss: 0.0187 - acc: 0.9955 - ETA: 1s - loss: 0.0184 - acc: 0.9956 - ETA: 1s - loss: 0.0183 - acc: 0.9958 - ETA: 1s - loss: 0.0182 - acc: 0.9959 - ETA: 1s - loss: 0.0183 - acc: 0.9958 - ETA: 1s - loss: 0.0187 - acc: 0.9954 - ETA: 1s - loss: 0.0183 - acc: 0.9955 - ETA: 1s - loss: 0.0181 - acc: 0.9957 - ETA: 1s - loss: 0.0185 - acc: 0.9953 - ETA: 1s - loss: 0.0181 - acc: 0.9955 - ETA: 1s - loss: 0.0179 - acc: 0.9956 - ETA: 1s - loss: 0.0178 - acc: 0.9957 - ETA: 0s - loss: 0.0175 - acc: 0.9958 - ETA: 0s - loss: 0.0176 - acc: 0.9957 - ETA: 0s - loss: 0.0175 - acc: 0.9958 - ETA: 0s - loss: 0.0174 - acc: 0.9959 - ETA: 0s - loss: 0.0174 - acc: 0.9958 - ETA: 0s - loss: 0.0182 - acc: 0.9955 - ETA: 0s - loss: 0.0183 - acc: 0.9956 - ETA: 0s - loss: 0.0181 - acc: 0.9957 - ETA: 0s - loss: 0.0184 - acc: 0.9956 - ETA: 0s - loss: 0.0188 - acc: 0.9955 - ETA: 0s - loss: 0.0197 - acc: 0.9955 - ETA: 0s - loss: 0.0196 - acc: 0.9954 - ETA: 0s - loss: 0.0194 - acc: 0.9955 - ETA: 0s - loss: 0.0204 - acc: 0.9953 - ETA: 0s - loss: 0.0212 - acc: 0.9950 - ETA: 0s - loss: 0.0214 - acc: 0.9948Epoch 00010: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0216 - acc: 0.9948 - val_loss: 0.7344 - val_acc: 0.8347
    Epoch 12/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.0022 - acc: 1.0000 - ETA: 2s - loss: 0.0067 - acc: 1.0000 - ETA: 2s - loss: 0.0065 - acc: 1.0000 - ETA: 2s - loss: 0.0053 - acc: 1.0000 - ETA: 2s - loss: 0.0059 - acc: 1.0000 - ETA: 2s - loss: 0.0055 - acc: 1.0000 - ETA: 2s - loss: 0.0061 - acc: 1.0000 - ETA: 2s - loss: 0.0074 - acc: 1.0000 - ETA: 2s - loss: 0.0076 - acc: 1.0000 - ETA: 2s - loss: 0.0071 - acc: 1.0000 - ETA: 2s - loss: 0.0086 - acc: 0.9992 - ETA: 2s - loss: 0.0083 - acc: 0.9992 - ETA: 2s - loss: 0.0079 - acc: 0.9993 - ETA: 2s - loss: 0.0077 - acc: 0.9994 - ETA: 2s - loss: 0.0092 - acc: 0.9988 - ETA: 2s - loss: 0.0088 - acc: 0.9989 - ETA: 2s - loss: 0.0086 - acc: 0.9989 - ETA: 2s - loss: 0.0084 - acc: 0.9990 - ETA: 2s - loss: 0.0085 - acc: 0.9986 - ETA: 2s - loss: 0.0083 - acc: 0.9986 - ETA: 2s - loss: 0.0083 - acc: 0.9987 - ETA: 2s - loss: 0.0083 - acc: 0.9987 - ETA: 1s - loss: 0.0081 - acc: 0.9988 - ETA: 1s - loss: 0.0085 - acc: 0.9985 - ETA: 1s - loss: 0.0087 - acc: 0.9981 - ETA: 1s - loss: 0.0085 - acc: 0.9982 - ETA: 1s - loss: 0.0083 - acc: 0.9983 - ETA: 1s - loss: 0.0083 - acc: 0.9983 - ETA: 1s - loss: 0.0084 - acc: 0.9984 - ETA: 1s - loss: 0.0083 - acc: 0.9984 - ETA: 1s - loss: 0.0083 - acc: 0.9985 - ETA: 1s - loss: 0.0088 - acc: 0.9982 - ETA: 1s - loss: 0.0112 - acc: 0.9977 - ETA: 1s - loss: 0.0111 - acc: 0.9978 - ETA: 1s - loss: 0.0124 - acc: 0.9976 - ETA: 1s - loss: 0.0122 - acc: 0.9976 - ETA: 1s - loss: 0.0123 - acc: 0.9977 - ETA: 1s - loss: 0.0131 - acc: 0.9975 - ETA: 1s - loss: 0.0130 - acc: 0.9976 - ETA: 1s - loss: 0.0128 - acc: 0.9976 - ETA: 1s - loss: 0.0126 - acc: 0.9977 - ETA: 1s - loss: 0.0124 - acc: 0.9977 - ETA: 1s - loss: 0.0124 - acc: 0.9978 - ETA: 1s - loss: 0.0128 - acc: 0.9974 - ETA: 0s - loss: 0.0127 - acc: 0.9974 - ETA: 0s - loss: 0.0126 - acc: 0.9975 - ETA: 0s - loss: 0.0129 - acc: 0.9973 - ETA: 0s - loss: 0.0132 - acc: 0.9972 - ETA: 0s - loss: 0.0130 - acc: 0.9973 - ETA: 0s - loss: 0.0129 - acc: 0.9973 - ETA: 0s - loss: 0.0128 - acc: 0.9974 - ETA: 0s - loss: 0.0127 - acc: 0.9974 - ETA: 0s - loss: 0.0127 - acc: 0.9975 - ETA: 0s - loss: 0.0132 - acc: 0.9970 - ETA: 0s - loss: 0.0149 - acc: 0.9965 - ETA: 0s - loss: 0.0147 - acc: 0.9966 - ETA: 0s - loss: 0.0146 - acc: 0.9966 - ETA: 0s - loss: 0.0153 - acc: 0.9965 - ETA: 0s - loss: 0.0156 - acc: 0.9964 - ETA: 0s - loss: 0.0162 - acc: 0.9963 - ETA: 0s - loss: 0.0178 - acc: 0.9959 - ETA: 0s - loss: 0.0176 - acc: 0.9959 - ETA: 0s - loss: 0.0178 - acc: 0.9958 - ETA: 0s - loss: 0.0177 - acc: 0.9959Epoch 00011: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0186 - acc: 0.9958 - val_loss: 0.7949 - val_acc: 0.8251
    Epoch 13/20
    6640/6680 [============================>.] - ETA: 3s - loss: 0.0142 - acc: 1.0000 - ETA: 3s - loss: 0.0166 - acc: 0.9917 - ETA: 3s - loss: 0.0101 - acc: 0.9955 - ETA: 3s - loss: 0.0079 - acc: 0.9969 - ETA: 3s - loss: 0.0091 - acc: 0.9952 - ETA: 3s - loss: 0.0077 - acc: 0.9962 - ETA: 3s - loss: 0.0075 - acc: 0.9968 - ETA: 3s - loss: 0.0075 - acc: 0.9972 - ETA: 2s - loss: 0.0069 - acc: 0.9976 - ETA: 2s - loss: 0.0067 - acc: 0.9978 - ETA: 2s - loss: 0.0064 - acc: 0.9980 - ETA: 2s - loss: 0.0062 - acc: 0.9982 - ETA: 2s - loss: 0.0060 - acc: 0.9984 - ETA: 2s - loss: 0.0060 - acc: 0.9986 - ETA: 2s - loss: 0.0065 - acc: 0.9987 - ETA: 2s - loss: 0.0079 - acc: 0.9975 - ETA: 2s - loss: 0.0095 - acc: 0.9971 - ETA: 2s - loss: 0.0095 - acc: 0.9968 - ETA: 2s - loss: 0.0099 - acc: 0.9965 - ETA: 2s - loss: 0.0104 - acc: 0.9962 - ETA: 2s - loss: 0.0104 - acc: 0.9964 - ETA: 2s - loss: 0.0105 - acc: 0.9961 - ETA: 2s - loss: 0.0112 - acc: 0.9959 - ETA: 1s - loss: 0.0108 - acc: 0.9961 - ETA: 1s - loss: 0.0106 - acc: 0.9963 - ETA: 1s - loss: 0.0103 - acc: 0.9964 - ETA: 1s - loss: 0.0104 - acc: 0.9966 - ETA: 1s - loss: 0.0103 - acc: 0.9967 - ETA: 1s - loss: 0.0103 - acc: 0.9968 - ETA: 1s - loss: 0.0100 - acc: 0.9970 - ETA: 1s - loss: 0.0097 - acc: 0.9971 - ETA: 1s - loss: 0.0096 - acc: 0.9972 - ETA: 1s - loss: 0.0094 - acc: 0.9973 - ETA: 1s - loss: 0.0092 - acc: 0.9973 - ETA: 1s - loss: 0.0090 - acc: 0.9974 - ETA: 1s - loss: 0.0089 - acc: 0.9975 - ETA: 1s - loss: 0.0091 - acc: 0.9973 - ETA: 1s - loss: 0.0089 - acc: 0.9974 - ETA: 1s - loss: 0.0091 - acc: 0.9972 - ETA: 1s - loss: 0.0095 - acc: 0.9971 - ETA: 0s - loss: 0.0095 - acc: 0.9972 - ETA: 0s - loss: 0.0095 - acc: 0.9972 - ETA: 0s - loss: 0.0099 - acc: 0.9971 - ETA: 0s - loss: 0.0098 - acc: 0.9972 - ETA: 0s - loss: 0.0102 - acc: 0.9970 - ETA: 0s - loss: 0.0105 - acc: 0.9967 - ETA: 0s - loss: 0.0107 - acc: 0.9966 - ETA: 0s - loss: 0.0106 - acc: 0.9967 - ETA: 0s - loss: 0.0107 - acc: 0.9968 - ETA: 0s - loss: 0.0107 - acc: 0.9967 - ETA: 0s - loss: 0.0106 - acc: 0.9967 - ETA: 0s - loss: 0.0105 - acc: 0.9968 - ETA: 0s - loss: 0.0104 - acc: 0.9969 - ETA: 0s - loss: 0.0106 - acc: 0.9968 - ETA: 0s - loss: 0.0114 - acc: 0.9967 - ETA: 0s - loss: 0.0117 - acc: 0.9966 - ETA: 0s - loss: 0.0119 - acc: 0.9965 - ETA: 0s - loss: 0.0120 - acc: 0.9964Epoch 00012: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0121 - acc: 0.9964 - val_loss: 0.8239 - val_acc: 0.8192
    Epoch 14/20
    6620/6680 [============================>.] - ETA: 3s - loss: 8.2649e-04 - acc: 1.0000 - ETA: 2s - loss: 0.0096 - acc: 0.9929     - ETA: 2s - loss: 0.0060 - acc: 0.9962 - ETA: 2s - loss: 0.0054 - acc: 0.9974 - ETA: 2s - loss: 0.0051 - acc: 0.9980 - ETA: 2s - loss: 0.0062 - acc: 0.9968 - ETA: 2s - loss: 0.0057 - acc: 0.9973 - ETA: 2s - loss: 0.0058 - acc: 0.9977 - ETA: 2s - loss: 0.0067 - acc: 0.9969 - ETA: 2s - loss: 0.0083 - acc: 0.9964 - ETA: 2s - loss: 0.0077 - acc: 0.9967 - ETA: 2s - loss: 0.0075 - acc: 0.9970 - ETA: 2s - loss: 0.0073 - acc: 0.9973 - ETA: 2s - loss: 0.0095 - acc: 0.9968 - ETA: 2s - loss: 0.0091 - acc: 0.9971 - ETA: 2s - loss: 0.0087 - acc: 0.9973 - ETA: 2s - loss: 0.0090 - acc: 0.9969 - ETA: 2s - loss: 0.0086 - acc: 0.9971 - ETA: 1s - loss: 0.0083 - acc: 0.9972 - ETA: 1s - loss: 0.0080 - acc: 0.9974 - ETA: 1s - loss: 0.0078 - acc: 0.9975 - ETA: 1s - loss: 0.0084 - acc: 0.9972 - ETA: 1s - loss: 0.0083 - acc: 0.9973 - ETA: 1s - loss: 0.0091 - acc: 0.9967 - ETA: 1s - loss: 0.0089 - acc: 0.9968 - ETA: 1s - loss: 0.0088 - acc: 0.9969 - ETA: 1s - loss: 0.0085 - acc: 0.9970 - ETA: 1s - loss: 0.0083 - acc: 0.9971 - ETA: 1s - loss: 0.0081 - acc: 0.9972 - ETA: 1s - loss: 0.0080 - acc: 0.9973 - ETA: 1s - loss: 0.0092 - acc: 0.9968 - ETA: 1s - loss: 0.0090 - acc: 0.9969 - ETA: 1s - loss: 0.0090 - acc: 0.9970 - ETA: 1s - loss: 0.0088 - acc: 0.9971 - ETA: 1s - loss: 0.0090 - acc: 0.9969 - ETA: 1s - loss: 0.0098 - acc: 0.9967 - ETA: 1s - loss: 0.0097 - acc: 0.9968 - ETA: 1s - loss: 0.0096 - acc: 0.9969 - ETA: 1s - loss: 0.0096 - acc: 0.9969 - ETA: 1s - loss: 0.0095 - acc: 0.9970 - ETA: 1s - loss: 0.0093 - acc: 0.9971 - ETA: 0s - loss: 0.0096 - acc: 0.9969 - ETA: 0s - loss: 0.0096 - acc: 0.9970 - ETA: 0s - loss: 0.0094 - acc: 0.9971 - ETA: 0s - loss: 0.0096 - acc: 0.9969 - ETA: 0s - loss: 0.0096 - acc: 0.9970 - ETA: 0s - loss: 0.0111 - acc: 0.9969 - ETA: 0s - loss: 0.0111 - acc: 0.9969 - ETA: 0s - loss: 0.0110 - acc: 0.9970 - ETA: 0s - loss: 0.0109 - acc: 0.9970 - ETA: 0s - loss: 0.0109 - acc: 0.9969 - ETA: 0s - loss: 0.0111 - acc: 0.9968 - ETA: 0s - loss: 0.0113 - acc: 0.9967 - ETA: 0s - loss: 0.0120 - acc: 0.9966 - ETA: 0s - loss: 0.0119 - acc: 0.9966 - ETA: 0s - loss: 0.0120 - acc: 0.9965 - ETA: 0s - loss: 0.0118 - acc: 0.9966 - ETA: 0s - loss: 0.0121 - acc: 0.9963 - ETA: 0s - loss: 0.0120 - acc: 0.9964 - ETA: 0s - loss: 0.0119 - acc: 0.9964 - ETA: 0s - loss: 0.0117 - acc: 0.9965 - ETA: 0s - loss: 0.0120 - acc: 0.9964Epoch 00013: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0119 - acc: 0.9964 - val_loss: 0.8153 - val_acc: 0.8335
    Epoch 15/20
    6560/6680 [============================>.] - ETA: 3s - loss: 0.0054 - acc: 1.0000 - ETA: 3s - loss: 0.0019 - acc: 1.0000 - ETA: 3s - loss: 0.0293 - acc: 0.9955 - ETA: 3s - loss: 0.0208 - acc: 0.9969 - ETA: 3s - loss: 0.0173 - acc: 0.9976 - ETA: 3s - loss: 0.0145 - acc: 0.9981 - ETA: 3s - loss: 0.0156 - acc: 0.9968 - ETA: 3s - loss: 0.0142 - acc: 0.9972 - ETA: 3s - loss: 0.0127 - acc: 0.9976 - ETA: 2s - loss: 0.0117 - acc: 0.9978 - ETA: 2s - loss: 0.0108 - acc: 0.9980 - ETA: 2s - loss: 0.0111 - acc: 0.9973 - ETA: 2s - loss: 0.0103 - acc: 0.9975 - ETA: 2s - loss: 0.0099 - acc: 0.9977 - ETA: 2s - loss: 0.0104 - acc: 0.9972 - ETA: 2s - loss: 0.0108 - acc: 0.9967 - ETA: 2s - loss: 0.0106 - acc: 0.9969 - ETA: 2s - loss: 0.0122 - acc: 0.9965 - ETA: 2s - loss: 0.0116 - acc: 0.9967 - ETA: 2s - loss: 0.0112 - acc: 0.9969 - ETA: 2s - loss: 0.0110 - acc: 0.9970 - ETA: 2s - loss: 0.0105 - acc: 0.9972 - ETA: 2s - loss: 0.0102 - acc: 0.9973 - ETA: 2s - loss: 0.0099 - acc: 0.9974 - ETA: 2s - loss: 0.0095 - acc: 0.9975 - ETA: 2s - loss: 0.0093 - acc: 0.9976 - ETA: 2s - loss: 0.0090 - acc: 0.9977 - ETA: 2s - loss: 0.0087 - acc: 0.9978 - ETA: 1s - loss: 0.0085 - acc: 0.9979 - ETA: 1s - loss: 0.0083 - acc: 0.9980 - ETA: 1s - loss: 0.0081 - acc: 0.9981 - ETA: 1s - loss: 0.0090 - acc: 0.9978 - ETA: 1s - loss: 0.0088 - acc: 0.9979 - ETA: 1s - loss: 0.0086 - acc: 0.9980 - ETA: 1s - loss: 0.0090 - acc: 0.9978 - ETA: 1s - loss: 0.0087 - acc: 0.9978 - ETA: 1s - loss: 0.0100 - acc: 0.9974 - ETA: 1s - loss: 0.0100 - acc: 0.9972 - ETA: 1s - loss: 0.0097 - acc: 0.9973 - ETA: 1s - loss: 0.0095 - acc: 0.9974 - ETA: 1s - loss: 0.0093 - acc: 0.9974 - ETA: 1s - loss: 0.0091 - acc: 0.9975 - ETA: 1s - loss: 0.0096 - acc: 0.9973 - ETA: 0s - loss: 0.0096 - acc: 0.9972 - ETA: 0s - loss: 0.0099 - acc: 0.9968 - ETA: 0s - loss: 0.0097 - acc: 0.9969 - ETA: 0s - loss: 0.0100 - acc: 0.9968 - ETA: 0s - loss: 0.0098 - acc: 0.9969 - ETA: 0s - loss: 0.0096 - acc: 0.9969 - ETA: 0s - loss: 0.0095 - acc: 0.9970 - ETA: 0s - loss: 0.0094 - acc: 0.9971 - ETA: 0s - loss: 0.0093 - acc: 0.9971 - ETA: 0s - loss: 0.0092 - acc: 0.9972 - ETA: 0s - loss: 0.0090 - acc: 0.9973 - ETA: 0s - loss: 0.0089 - acc: 0.9973 - ETA: 0s - loss: 0.0088 - acc: 0.9974 - ETA: 0s - loss: 0.0087 - acc: 0.9974 - ETA: 0s - loss: 0.0087 - acc: 0.9975 - ETA: 0s - loss: 0.0086 - acc: 0.9975 - ETA: 0s - loss: 0.0086 - acc: 0.9976Epoch 00014: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0087 - acc: 0.9976 - val_loss: 0.8397 - val_acc: 0.8275
    Epoch 16/20
    6600/6680 [============================>.] - ETA: 3s - loss: 0.0012 - acc: 1.0000 - ETA: 2s - loss: 0.0243 - acc: 0.9929 - ETA: 2s - loss: 0.0135 - acc: 0.9962 - ETA: 2s - loss: 0.0098 - acc: 0.9974 - ETA: 2s - loss: 0.0142 - acc: 0.9940 - ETA: 2s - loss: 0.0128 - acc: 0.9952 - ETA: 2s - loss: 0.0109 - acc: 0.9959 - ETA: 2s - loss: 0.0095 - acc: 0.9965 - ETA: 2s - loss: 0.0086 - acc: 0.9969 - ETA: 2s - loss: 0.0078 - acc: 0.9973 - ETA: 2s - loss: 0.0071 - acc: 0.9975 - ETA: 2s - loss: 0.0077 - acc: 0.9970 - ETA: 2s - loss: 0.0074 - acc: 0.9973 - ETA: 2s - loss: 0.0070 - acc: 0.9975 - ETA: 2s - loss: 0.0067 - acc: 0.9976 - ETA: 2s - loss: 0.0063 - acc: 0.9978 - ETA: 2s - loss: 0.0061 - acc: 0.9979 - ETA: 2s - loss: 0.0074 - acc: 0.9976 - ETA: 1s - loss: 0.0071 - acc: 0.9977 - ETA: 1s - loss: 0.0109 - acc: 0.9974 - ETA: 1s - loss: 0.0105 - acc: 0.9975 - ETA: 1s - loss: 0.0106 - acc: 0.9972 - ETA: 1s - loss: 0.0101 - acc: 0.9974 - ETA: 1s - loss: 0.0098 - acc: 0.9975 - ETA: 1s - loss: 0.0094 - acc: 0.9976 - ETA: 1s - loss: 0.0093 - acc: 0.9977 - ETA: 1s - loss: 0.0092 - acc: 0.9978 - ETA: 1s - loss: 0.0090 - acc: 0.9979 - ETA: 1s - loss: 0.0087 - acc: 0.9979 - ETA: 1s - loss: 0.0085 - acc: 0.9980 - ETA: 1s - loss: 0.0085 - acc: 0.9981 - ETA: 1s - loss: 0.0084 - acc: 0.9981 - ETA: 1s - loss: 0.0081 - acc: 0.9982 - ETA: 1s - loss: 0.0079 - acc: 0.9982 - ETA: 1s - loss: 0.0083 - acc: 0.9980 - ETA: 1s - loss: 0.0085 - acc: 0.9979 - ETA: 1s - loss: 0.0084 - acc: 0.9979 - ETA: 0s - loss: 0.0086 - acc: 0.9978 - ETA: 0s - loss: 0.0084 - acc: 0.9978 - ETA: 0s - loss: 0.0082 - acc: 0.9979 - ETA: 0s - loss: 0.0089 - acc: 0.9977 - ETA: 0s - loss: 0.0087 - acc: 0.9978 - ETA: 0s - loss: 0.0087 - acc: 0.9978 - ETA: 0s - loss: 0.0088 - acc: 0.9977 - ETA: 0s - loss: 0.0087 - acc: 0.9977 - ETA: 0s - loss: 0.0088 - acc: 0.9976 - ETA: 0s - loss: 0.0090 - acc: 0.9975 - ETA: 0s - loss: 0.0088 - acc: 0.9975 - ETA: 0s - loss: 0.0089 - acc: 0.9974 - ETA: 0s - loss: 0.0088 - acc: 0.9974 - ETA: 0s - loss: 0.0087 - acc: 0.9975 - ETA: 0s - loss: 0.0086 - acc: 0.9975 - ETA: 0s - loss: 0.0084 - acc: 0.9976 - ETA: 0s - loss: 0.0083 - acc: 0.9976 - ETA: 0s - loss: 0.0082 - acc: 0.9977 - ETA: 0s - loss: 0.0081 - acc: 0.9977Epoch 00015: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0086 - acc: 0.9975 - val_loss: 0.8601 - val_acc: 0.8347
    Epoch 17/20
    6600/6680 [============================>.] - ETA: 3s - loss: 2.1699e-04 - acc: 1.0000 - ETA: 2s - loss: 8.0555e-04 - acc: 1.0000 - ETA: 2s - loss: 8.0681e-04 - acc: 1.0000 - ETA: 2s - loss: 0.0101 - acc: 0.9974     - ETA: 2s - loss: 0.0078 - acc: 0.9980 - ETA: 2s - loss: 0.0068 - acc: 0.9984 - ETA: 2s - loss: 0.0062 - acc: 0.9986 - ETA: 2s - loss: 0.0060 - acc: 0.9988 - ETA: 2s - loss: 0.0054 - acc: 0.9989 - ETA: 2s - loss: 0.0051 - acc: 0.9990 - ETA: 2s - loss: 0.0071 - acc: 0.9982 - ETA: 2s - loss: 0.0066 - acc: 0.9984 - ETA: 2s - loss: 0.0062 - acc: 0.9985 - ETA: 2s - loss: 0.0061 - acc: 0.9986 - ETA: 2s - loss: 0.0057 - acc: 0.9987 - ETA: 2s - loss: 0.0055 - acc: 0.9988 - ETA: 2s - loss: 0.0052 - acc: 0.9988 - ETA: 2s - loss: 0.0050 - acc: 0.9989 - ETA: 2s - loss: 0.0047 - acc: 0.9990 - ETA: 2s - loss: 0.0046 - acc: 0.9990 - ETA: 2s - loss: 0.0044 - acc: 0.9991 - ETA: 2s - loss: 0.0044 - acc: 0.9991 - ETA: 2s - loss: 0.0043 - acc: 0.9991 - ETA: 2s - loss: 0.0041 - acc: 0.9992 - ETA: 2s - loss: 0.0040 - acc: 0.9992 - ETA: 1s - loss: 0.0044 - acc: 0.9989 - ETA: 1s - loss: 0.0057 - acc: 0.9985 - ETA: 1s - loss: 0.0055 - acc: 0.9986 - ETA: 1s - loss: 0.0054 - acc: 0.9986 - ETA: 1s - loss: 0.0052 - acc: 0.9987 - ETA: 1s - loss: 0.0051 - acc: 0.9987 - ETA: 1s - loss: 0.0052 - acc: 0.9985 - ETA: 1s - loss: 0.0051 - acc: 0.9985 - ETA: 1s - loss: 0.0051 - acc: 0.9985 - ETA: 1s - loss: 0.0050 - acc: 0.9986 - ETA: 1s - loss: 0.0072 - acc: 0.9984 - ETA: 1s - loss: 0.0074 - acc: 0.9981 - ETA: 1s - loss: 0.0072 - acc: 0.9982 - ETA: 1s - loss: 0.0071 - acc: 0.9982 - ETA: 1s - loss: 0.0077 - acc: 0.9980 - ETA: 1s - loss: 0.0076 - acc: 0.9981 - ETA: 1s - loss: 0.0075 - acc: 0.9981 - ETA: 1s - loss: 0.0074 - acc: 0.9982 - ETA: 1s - loss: 0.0073 - acc: 0.9982 - ETA: 1s - loss: 0.0081 - acc: 0.9980 - ETA: 1s - loss: 0.0081 - acc: 0.9981 - ETA: 0s - loss: 0.0084 - acc: 0.9979 - ETA: 0s - loss: 0.0083 - acc: 0.9979 - ETA: 0s - loss: 0.0086 - acc: 0.9976 - ETA: 0s - loss: 0.0085 - acc: 0.9976 - ETA: 0s - loss: 0.0083 - acc: 0.9977 - ETA: 0s - loss: 0.0082 - acc: 0.9977 - ETA: 0s - loss: 0.0081 - acc: 0.9978 - ETA: 0s - loss: 0.0081 - acc: 0.9978 - ETA: 0s - loss: 0.0080 - acc: 0.9978 - ETA: 0s - loss: 0.0079 - acc: 0.9979 - ETA: 0s - loss: 0.0078 - acc: 0.9979 - ETA: 0s - loss: 0.0078 - acc: 0.9980 - ETA: 0s - loss: 0.0077 - acc: 0.9980 - ETA: 0s - loss: 0.0076 - acc: 0.9980 - ETA: 0s - loss: 0.0075 - acc: 0.9981 - ETA: 0s - loss: 0.0075 - acc: 0.9981 - ETA: 0s - loss: 0.0074 - acc: 0.9981 - ETA: 0s - loss: 0.0073 - acc: 0.9982 - ETA: 0s - loss: 0.0073 - acc: 0.9982Epoch 00016: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0072 - acc: 0.9982 - val_loss: 0.9362 - val_acc: 0.8275
    Epoch 18/20
    6620/6680 [============================>.] - ETA: 3s - loss: 0.0177 - acc: 1.0000 - ETA: 3s - loss: 0.0091 - acc: 1.0000 - ETA: 3s - loss: 0.0054 - acc: 1.0000 - ETA: 3s - loss: 0.0040 - acc: 1.0000 - ETA: 3s - loss: 0.0145 - acc: 0.9977 - ETA: 3s - loss: 0.0119 - acc: 0.9981 - ETA: 3s - loss: 0.0102 - acc: 0.9984 - ETA: 3s - loss: 0.0088 - acc: 0.9986 - ETA: 2s - loss: 0.0079 - acc: 0.9988 - ETA: 2s - loss: 0.0072 - acc: 0.9989 - ETA: 2s - loss: 0.0086 - acc: 0.9981 - ETA: 2s - loss: 0.0080 - acc: 0.9982 - ETA: 2s - loss: 0.0074 - acc: 0.9984 - ETA: 2s - loss: 0.0069 - acc: 0.9985 - ETA: 2s - loss: 0.0065 - acc: 0.9986 - ETA: 2s - loss: 0.0065 - acc: 0.9987 - ETA: 2s - loss: 0.0062 - acc: 0.9988 - ETA: 2s - loss: 0.0078 - acc: 0.9983 - ETA: 2s - loss: 0.0086 - acc: 0.9978 - ETA: 2s - loss: 0.0082 - acc: 0.9980 - ETA: 2s - loss: 0.0079 - acc: 0.9981 - ETA: 2s - loss: 0.0077 - acc: 0.9982 - ETA: 2s - loss: 0.0078 - acc: 0.9978 - ETA: 2s - loss: 0.0076 - acc: 0.9979 - ETA: 2s - loss: 0.0073 - acc: 0.9980 - ETA: 1s - loss: 0.0070 - acc: 0.9981 - ETA: 1s - loss: 0.0069 - acc: 0.9982 - ETA: 1s - loss: 0.0084 - acc: 0.9979 - ETA: 1s - loss: 0.0082 - acc: 0.9980 - ETA: 1s - loss: 0.0079 - acc: 0.9981 - ETA: 1s - loss: 0.0083 - acc: 0.9979 - ETA: 1s - loss: 0.0081 - acc: 0.9979 - ETA: 1s - loss: 0.0078 - acc: 0.9980 - ETA: 1s - loss: 0.0077 - acc: 0.9981 - ETA: 1s - loss: 0.0076 - acc: 0.9981 - ETA: 1s - loss: 0.0074 - acc: 0.9982 - ETA: 1s - loss: 0.0072 - acc: 0.9982 - ETA: 1s - loss: 0.0071 - acc: 0.9983 - ETA: 1s - loss: 0.0070 - acc: 0.9983 - ETA: 1s - loss: 0.0069 - acc: 0.9984 - ETA: 1s - loss: 0.0067 - acc: 0.9984 - ETA: 0s - loss: 0.0075 - acc: 0.9983 - ETA: 0s - loss: 0.0073 - acc: 0.9983 - ETA: 0s - loss: 0.0072 - acc: 0.9983 - ETA: 0s - loss: 0.0071 - acc: 0.9984 - ETA: 0s - loss: 0.0071 - acc: 0.9984 - ETA: 0s - loss: 0.0069 - acc: 0.9985 - ETA: 0s - loss: 0.0068 - acc: 0.9985 - ETA: 0s - loss: 0.0067 - acc: 0.9985 - ETA: 0s - loss: 0.0066 - acc: 0.9986 - ETA: 0s - loss: 0.0064 - acc: 0.9986 - ETA: 0s - loss: 0.0063 - acc: 0.9986 - ETA: 0s - loss: 0.0063 - acc: 0.9986 - ETA: 0s - loss: 0.0062 - acc: 0.9987 - ETA: 0s - loss: 0.0061 - acc: 0.9987 - ETA: 0s - loss: 0.0060 - acc: 0.9987 - ETA: 0s - loss: 0.0060 - acc: 0.9987 - ETA: 0s - loss: 0.0060 - acc: 0.9988 - ETA: 0s - loss: 0.0059 - acc: 0.9988Epoch 00017: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0058 - acc: 0.9988 - val_loss: 0.8732 - val_acc: 0.8228
    Epoch 19/20
    6620/6680 [============================>.] - ETA: 3s - loss: 1.1440e-04 - acc: 1.0000 - ETA: 2s - loss: 4.2132e-04 - acc: 1.0000 - ETA: 2s - loss: 4.2250e-04 - acc: 1.0000 - ETA: 2s - loss: 4.2764e-04 - acc: 1.0000 - ETA: 2s - loss: 4.1593e-04 - acc: 1.0000 - ETA: 2s - loss: 4.1321e-04 - acc: 1.0000 - ETA: 2s - loss: 3.7705e-04 - acc: 1.0000 - ETA: 2s - loss: 3.6876e-04 - acc: 1.0000 - ETA: 2s - loss: 3.6417e-04 - acc: 1.0000 - ETA: 2s - loss: 4.5136e-04 - acc: 1.0000 - ETA: 2s - loss: 4.6121e-04 - acc: 1.0000 - ETA: 2s - loss: 4.8916e-04 - acc: 1.0000 - ETA: 2s - loss: 4.8452e-04 - acc: 1.0000 - ETA: 2s - loss: 5.1069e-04 - acc: 1.0000 - ETA: 2s - loss: 5.5216e-04 - acc: 1.0000 - ETA: 2s - loss: 5.8183e-04 - acc: 1.0000 - ETA: 2s - loss: 6.3211e-04 - acc: 1.0000 - ETA: 2s - loss: 6.3338e-04 - acc: 1.0000 - ETA: 2s - loss: 6.8819e-04 - acc: 1.0000 - ETA: 1s - loss: 0.0024 - acc: 0.9996     - ETA: 1s - loss: 0.0024 - acc: 0.9996 - ETA: 1s - loss: 0.0023 - acc: 0.9996 - ETA: 1s - loss: 0.0028 - acc: 0.9992 - ETA: 1s - loss: 0.0027 - acc: 0.9993 - ETA: 1s - loss: 0.0030 - acc: 0.9990 - ETA: 1s - loss: 0.0030 - acc: 0.9990 - ETA: 1s - loss: 0.0035 - acc: 0.9984 - ETA: 1s - loss: 0.0034 - acc: 0.9985 - ETA: 1s - loss: 0.0038 - acc: 0.9982 - ETA: 1s - loss: 0.0038 - acc: 0.9983 - ETA: 1s - loss: 0.0037 - acc: 0.9983 - ETA: 1s - loss: 0.0037 - acc: 0.9984 - ETA: 1s - loss: 0.0036 - acc: 0.9984 - ETA: 1s - loss: 0.0035 - acc: 0.9985 - ETA: 1s - loss: 0.0034 - acc: 0.9985 - ETA: 1s - loss: 0.0034 - acc: 0.9986 - ETA: 1s - loss: 0.0033 - acc: 0.9986 - ETA: 0s - loss: 0.0033 - acc: 0.9987 - ETA: 0s - loss: 0.0032 - acc: 0.9987 - ETA: 0s - loss: 0.0033 - acc: 0.9985 - ETA: 0s - loss: 0.0038 - acc: 0.9983 - ETA: 0s - loss: 0.0037 - acc: 0.9984 - ETA: 0s - loss: 0.0037 - acc: 0.9984 - ETA: 0s - loss: 0.0036 - acc: 0.9985 - ETA: 0s - loss: 0.0060 - acc: 0.9981 - ETA: 0s - loss: 0.0059 - acc: 0.9982 - ETA: 0s - loss: 0.0058 - acc: 0.9982 - ETA: 0s - loss: 0.0057 - acc: 0.9982 - ETA: 0s - loss: 0.0056 - acc: 0.9983 - ETA: 0s - loss: 0.0055 - acc: 0.9983 - ETA: 0s - loss: 0.0055 - acc: 0.9983 - ETA: 0s - loss: 0.0054 - acc: 0.9984 - ETA: 0s - loss: 0.0053 - acc: 0.9984 - ETA: 0s - loss: 0.0053 - acc: 0.9984 - ETA: 0s - loss: 0.0053 - acc: 0.9985 - ETA: 0s - loss: 0.0056 - acc: 0.9983Epoch 00018: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0056 - acc: 0.9984 - val_loss: 0.8781 - val_acc: 0.8275
    Epoch 20/20
    6620/6680 [============================>.] - ETA: 3s - loss: 0.0011 - acc: 1.0000 - ETA: 2s - loss: 0.0025 - acc: 1.0000 - ETA: 2s - loss: 0.0015 - acc: 1.0000 - ETA: 2s - loss: 0.0014 - acc: 1.0000 - ETA: 2s - loss: 0.0011 - acc: 1.0000 - ETA: 2s - loss: 9.4173e-04 - acc: 1.0000 - ETA: 2s - loss: 0.0043 - acc: 0.9986     - ETA: 2s - loss: 0.0038 - acc: 0.9988 - ETA: 2s - loss: 0.0034 - acc: 0.9990 - ETA: 2s - loss: 0.0031 - acc: 0.9991 - ETA: 2s - loss: 0.0085 - acc: 0.9984 - ETA: 2s - loss: 0.0078 - acc: 0.9985 - ETA: 2s - loss: 0.0074 - acc: 0.9986 - ETA: 2s - loss: 0.0069 - acc: 0.9987 - ETA: 2s - loss: 0.0064 - acc: 0.9988 - ETA: 2s - loss: 0.0060 - acc: 0.9989 - ETA: 2s - loss: 0.0056 - acc: 0.9990 - ETA: 2s - loss: 0.0054 - acc: 0.9990 - ETA: 1s - loss: 0.0051 - acc: 0.9991 - ETA: 1s - loss: 0.0057 - acc: 0.9987 - ETA: 1s - loss: 0.0054 - acc: 0.9988 - ETA: 1s - loss: 0.0054 - acc: 0.9988 - ETA: 1s - loss: 0.0052 - acc: 0.9989 - ETA: 1s - loss: 0.0050 - acc: 0.9989 - ETA: 1s - loss: 0.0049 - acc: 0.9990 - ETA: 1s - loss: 0.0047 - acc: 0.9990 - ETA: 1s - loss: 0.0046 - acc: 0.9990 - ETA: 1s - loss: 0.0044 - acc: 0.9991 - ETA: 1s - loss: 0.0060 - acc: 0.9988 - ETA: 1s - loss: 0.0074 - acc: 0.9983 - ETA: 1s - loss: 0.0072 - acc: 0.9983 - ETA: 1s - loss: 0.0069 - acc: 0.9984 - ETA: 1s - loss: 0.0075 - acc: 0.9982 - ETA: 1s - loss: 0.0073 - acc: 0.9982 - ETA: 1s - loss: 0.0072 - acc: 0.9983 - ETA: 1s - loss: 0.0070 - acc: 0.9983 - ETA: 1s - loss: 0.0068 - acc: 0.9984 - ETA: 0s - loss: 0.0071 - acc: 0.9982 - ETA: 0s - loss: 0.0070 - acc: 0.9983 - ETA: 0s - loss: 0.0068 - acc: 0.9983 - ETA: 0s - loss: 0.0066 - acc: 0.9983 - ETA: 0s - loss: 0.0068 - acc: 0.9982 - ETA: 0s - loss: 0.0077 - acc: 0.9978 - ETA: 0s - loss: 0.0075 - acc: 0.9979 - ETA: 0s - loss: 0.0074 - acc: 0.9979 - ETA: 0s - loss: 0.0072 - acc: 0.9980 - ETA: 0s - loss: 0.0071 - acc: 0.9980 - ETA: 0s - loss: 0.0069 - acc: 0.9981 - ETA: 0s - loss: 0.0068 - acc: 0.9981 - ETA: 0s - loss: 0.0067 - acc: 0.9981 - ETA: 0s - loss: 0.0066 - acc: 0.9982 - ETA: 0s - loss: 0.0065 - acc: 0.9982 - ETA: 0s - loss: 0.0064 - acc: 0.9982 - ETA: 0s - loss: 0.0063 - acc: 0.9983 - ETA: 0s - loss: 0.0066 - acc: 0.9982 - ETA: 0s - loss: 0.0065 - acc: 0.9982Epoch 00019: val_loss did not improve
    6680/6680 [==============================] - 3s - loss: 0.0064 - acc: 0.9982 - val_loss: 0.9351 - val_acc: 0.8335
    ---I am done saving model valid_Resnet50 ----
    
    In [40]:
    ### TODO: Train the model.
    checkpointer_InceptionV3 = ModelCheckpoint(filepath='weights.best.InceptionV3.hdf5', 
                                   verbose=1, save_best_only=True)
    
    InceptionV3_model.fit(train_InceptionV3, train_targets, 
              validation_data=(valid_InceptionV3, valid_targets),
              epochs=20, batch_size=20, callbacks=[checkpointer_InceptionV3], verbose=1)
    
    print('---I am done saving model valid_InceptionV3 ----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/20
    6660/6680 [============================>.] - ETA: 667s - loss: 5.0949 - acc: 0.0000e+00 - ETA: 485s - loss: 6.2580 - acc: 0.0250     - ETA: 382s - loss: 6.1590 - acc: 0.0167 - ETA: 330s - loss: 6.2435 - acc: 0.0125 - ETA: 279s - loss: 6.0845 - acc: 0.0200 - ETA: 241s - loss: 5.9221 - acc: 0.0333 - ETA: 210s - loss: 5.8357 - acc: 0.0357 - ETA: 186s - loss: 5.7328 - acc: 0.0625 - ETA: 170s - loss: 5.6272 - acc: 0.0722 - ETA: 155s - loss: 5.4601 - acc: 0.0900 - ETA: 144s - loss: 5.3829 - acc: 0.1045 - ETA: 134s - loss: 5.2416 - acc: 0.1250 - ETA: 125s - loss: 5.1214 - acc: 0.1346 - ETA: 119s - loss: 5.0546 - acc: 0.1393 - ETA: 113s - loss: 5.0031 - acc: 0.1467 - ETA: 108s - loss: 4.9075 - acc: 0.1656 - ETA: 103s - loss: 4.8837 - acc: 0.1706 - ETA: 100s - loss: 4.7967 - acc: 0.1833 - ETA: 97s - loss: 4.7396 - acc: 0.1921  - ETA: 94s - loss: 4.6796 - acc: 0.1950 - ETA: 92s - loss: 4.5828 - acc: 0.2048 - ETA: 88s - loss: 4.4852 - acc: 0.2227 - ETA: 85s - loss: 4.4295 - acc: 0.2304 - ETA: 83s - loss: 4.3600 - acc: 0.2458 - ETA: 81s - loss: 4.3151 - acc: 0.2520 - ETA: 78s - loss: 4.2523 - acc: 0.2596 - ETA: 76s - loss: 4.1595 - acc: 0.2741 - ETA: 74s - loss: 4.1082 - acc: 0.2786 - ETA: 72s - loss: 4.0460 - acc: 0.2845 - ETA: 70s - loss: 4.0030 - acc: 0.2900 - ETA: 69s - loss: 3.9498 - acc: 0.2952 - ETA: 68s - loss: 3.8972 - acc: 0.2984 - ETA: 66s - loss: 3.8453 - acc: 0.3076 - ETA: 65s - loss: 3.7864 - acc: 0.3162 - ETA: 64s - loss: 3.7340 - acc: 0.3200 - ETA: 62s - loss: 3.6760 - acc: 0.3292 - ETA: 61s - loss: 3.6246 - acc: 0.3351 - ETA: 60s - loss: 3.5750 - acc: 0.3382 - ETA: 59s - loss: 3.5414 - acc: 0.3397 - ETA: 58s - loss: 3.5080 - acc: 0.3425 - ETA: 57s - loss: 3.4688 - acc: 0.3500 - ETA: 56s - loss: 3.4178 - acc: 0.3583 - ETA: 55s - loss: 3.3757 - acc: 0.3628 - ETA: 54s - loss: 3.3275 - acc: 0.3682 - ETA: 53s - loss: 3.2955 - acc: 0.3711 - ETA: 52s - loss: 3.2575 - acc: 0.3750 - ETA: 52s - loss: 3.2243 - acc: 0.3809 - ETA: 51s - loss: 3.1888 - acc: 0.3865 - ETA: 50s - loss: 3.1607 - acc: 0.3929 - ETA: 49s - loss: 3.1230 - acc: 0.3970 - ETA: 49s - loss: 3.0855 - acc: 0.4020 - ETA: 48s - loss: 3.0737 - acc: 0.4019 - ETA: 47s - loss: 3.0535 - acc: 0.4038 - ETA: 47s - loss: 3.0246 - acc: 0.4065 - ETA: 46s - loss: 2.9849 - acc: 0.4136 - ETA: 46s - loss: 2.9521 - acc: 0.4196 - ETA: 45s - loss: 2.9252 - acc: 0.4211 - ETA: 44s - loss: 2.9015 - acc: 0.4233 - ETA: 44s - loss: 2.8707 - acc: 0.4263 - ETA: 43s - loss: 2.8451 - acc: 0.4300 - ETA: 43s - loss: 2.8086 - acc: 0.4377 - ETA: 43s - loss: 2.7797 - acc: 0.4411 - ETA: 42s - loss: 2.7570 - acc: 0.4452 - ETA: 42s - loss: 2.7310 - acc: 0.4492 - ETA: 41s - loss: 2.7121 - acc: 0.4515 - ETA: 41s - loss: 2.6853 - acc: 0.4568 - ETA: 41s - loss: 2.6609 - acc: 0.4597 - ETA: 40s - loss: 2.6412 - acc: 0.4640 - ETA: 40s - loss: 2.6257 - acc: 0.4659 - ETA: 40s - loss: 2.6082 - acc: 0.4693 - ETA: 39s - loss: 2.5971 - acc: 0.4704 - ETA: 39s - loss: 2.5793 - acc: 0.4743 - ETA: 39s - loss: 2.5646 - acc: 0.4740 - ETA: 38s - loss: 2.5408 - acc: 0.4791 - ETA: 39s - loss: 2.5159 - acc: 0.4833 - ETA: 38s - loss: 2.5002 - acc: 0.4849 - ETA: 38s - loss: 2.4748 - acc: 0.4896 - ETA: 37s - loss: 2.4612 - acc: 0.4910 - ETA: 37s - loss: 2.4269 - acc: 0.4981 - ETA: 36s - loss: 2.4071 - acc: 0.5019 - ETA: 36s - loss: 2.3897 - acc: 0.5030 - ETA: 36s - loss: 2.3763 - acc: 0.5054 - ETA: 35s - loss: 2.3626 - acc: 0.5071 - ETA: 35s - loss: 2.3486 - acc: 0.5094 - ETA: 35s - loss: 2.3276 - acc: 0.5134 - ETA: 35s - loss: 2.3190 - acc: 0.5144 - ETA: 34s - loss: 2.3054 - acc: 0.5170 - ETA: 34s - loss: 2.2974 - acc: 0.5169 - ETA: 34s - loss: 2.2815 - acc: 0.5194 - ETA: 34s - loss: 2.2676 - acc: 0.5214 - ETA: 33s - loss: 2.2491 - acc: 0.5250 - ETA: 33s - loss: 2.2378 - acc: 0.5269 - ETA: 33s - loss: 2.2240 - acc: 0.5282 - ETA: 33s - loss: 2.2080 - acc: 0.5316 - ETA: 32s - loss: 2.1944 - acc: 0.5344 - ETA: 32s - loss: 2.1801 - acc: 0.5361 - ETA: 32s - loss: 2.1690 - acc: 0.5378 - ETA: 31s - loss: 2.1582 - acc: 0.5389 - ETA: 31s - loss: 2.1466 - acc: 0.5405 - ETA: 31s - loss: 2.1365 - acc: 0.5411 - ETA: 31s - loss: 2.1238 - acc: 0.5431 - ETA: 31s - loss: 2.1164 - acc: 0.5437 - ETA: 31s - loss: 2.1079 - acc: 0.5442 - ETA: 31s - loss: 2.0996 - acc: 0.5443 - ETA: 31s - loss: 2.0894 - acc: 0.5458 - ETA: 30s - loss: 2.0793 - acc: 0.5477 - ETA: 30s - loss: 2.0686 - acc: 0.5491 - ETA: 30s - loss: 2.0562 - acc: 0.5514 - ETA: 30s - loss: 2.0465 - acc: 0.5532 - ETA: 30s - loss: 2.0331 - acc: 0.5563 - ETA: 29s - loss: 2.0214 - acc: 0.5589 - ETA: 29s - loss: 2.0105 - acc: 0.5611 - ETA: 29s - loss: 2.0080 - acc: 0.5610 - ETA: 29s - loss: 1.9990 - acc: 0.5622 - ETA: 29s - loss: 1.9879 - acc: 0.5642 - ETA: 29s - loss: 1.9776 - acc: 0.5654 - ETA: 28s - loss: 1.9683 - acc: 0.5657 - ETA: 28s - loss: 1.9592 - acc: 0.5660 - ETA: 28s - loss: 1.9519 - acc: 0.5671 - ETA: 28s - loss: 1.9419 - acc: 0.5686 - ETA: 28s - loss: 1.9304 - acc: 0.5705 - ETA: 28s - loss: 1.9217 - acc: 0.5720 - ETA: 27s - loss: 1.9123 - acc: 0.5738 - ETA: 27s - loss: 1.9002 - acc: 0.5768 - ETA: 27s - loss: 1.8933 - acc: 0.5778 - ETA: 27s - loss: 1.8817 - acc: 0.5799 - ETA: 27s - loss: 1.8746 - acc: 0.5805 - ETA: 26s - loss: 1.8680 - acc: 0.5814 - ETA: 26s - loss: 1.8586 - acc: 0.5827 - ETA: 26s - loss: 1.8496 - acc: 0.5844 - ETA: 26s - loss: 1.8437 - acc: 0.5852 - ETA: 26s - loss: 1.8377 - acc: 0.5861 - ETA: 25s - loss: 1.8283 - acc: 0.5873 - ETA: 25s - loss: 1.8195 - acc: 0.5885 - ETA: 25s - loss: 1.8098 - acc: 0.5897 - ETA: 25s - loss: 1.8019 - acc: 0.5916 - ETA: 25s - loss: 1.7958 - acc: 0.5913 - ETA: 25s - loss: 1.7897 - acc: 0.5928 - ETA: 25s - loss: 1.7866 - acc: 0.5936 - ETA: 24s - loss: 1.7794 - acc: 0.5947 - ETA: 24s - loss: 1.7711 - acc: 0.5961 - ETA: 24s - loss: 1.7680 - acc: 0.5962 - ETA: 24s - loss: 1.7614 - acc: 0.5976 - ETA: 24s - loss: 1.7521 - acc: 0.5993 - ETA: 24s - loss: 1.7447 - acc: 0.6003 - ETA: 23s - loss: 1.7363 - acc: 0.6024 - ETA: 23s - loss: 1.7261 - acc: 0.6044 - ETA: 23s - loss: 1.7201 - acc: 0.6060 - ETA: 23s - loss: 1.7162 - acc: 0.6070 - ETA: 23s - loss: 1.7076 - acc: 0.6086 - ETA: 23s - loss: 1.7032 - acc: 0.6095 - ETA: 22s - loss: 1.7024 - acc: 0.6098 - ETA: 22s - loss: 1.6978 - acc: 0.6104 - ETA: 22s - loss: 1.6902 - acc: 0.6119 - ETA: 22s - loss: 1.6870 - acc: 0.6128 - ETA: 22s - loss: 1.6801 - acc: 0.6140 - ETA: 21s - loss: 1.6722 - acc: 0.6155 - ETA: 21s - loss: 1.6664 - acc: 0.6167 - ETA: 21s - loss: 1.6637 - acc: 0.6175 - ETA: 21s - loss: 1.6603 - acc: 0.6171 - ETA: 21s - loss: 1.6534 - acc: 0.6191 - ETA: 21s - loss: 1.6532 - acc: 0.6193 - ETA: 21s - loss: 1.6468 - acc: 0.6204 - ETA: 21s - loss: 1.6390 - acc: 0.6218 - ETA: 20s - loss: 1.6327 - acc: 0.6229 - ETA: 20s - loss: 1.6268 - acc: 0.6240 - ETA: 20s - loss: 1.6228 - acc: 0.6241 - ETA: 20s - loss: 1.6179 - acc: 0.6251 - ETA: 20s - loss: 1.6122 - acc: 0.6265 - ETA: 19s - loss: 1.6011 - acc: 0.6282 - ETA: 19s - loss: 1.5952 - acc: 0.6295 - ETA: 19s - loss: 1.5911 - acc: 0.6302 - ETA: 19s - loss: 1.5855 - acc: 0.6311 - ETA: 19s - loss: 1.5747 - acc: 0.6331 - ETA: 19s - loss: 1.5685 - acc: 0.6340 - ETA: 19s - loss: 1.5640 - acc: 0.6352 - ETA: 19s - loss: 1.5596 - acc: 0.6358 - ETA: 19s - loss: 1.5566 - acc: 0.6365 - ETA: 19s - loss: 1.5530 - acc: 0.6365 - ETA: 19s - loss: 1.5496 - acc: 0.6369 - ETA: 19s - loss: 1.5446 - acc: 0.6380 - ETA: 19s - loss: 1.5397 - acc: 0.6384 - ETA: 19s - loss: 1.5344 - acc: 0.6395 - ETA: 19s - loss: 1.5304 - acc: 0.6404 - ETA: 19s - loss: 1.5247 - acc: 0.6418 - ETA: 19s - loss: 1.5198 - acc: 0.6426 - ETA: 19s - loss: 1.5141 - acc: 0.6437 - ETA: 19s - loss: 1.5114 - acc: 0.6445 - ETA: 19s - loss: 1.5086 - acc: 0.6453 - ETA: 19s - loss: 1.5071 - acc: 0.6453 - ETA: 19s - loss: 1.5019 - acc: 0.6459 - ETA: 18s - loss: 1.4963 - acc: 0.6469 - ETA: 18s - loss: 1.4921 - acc: 0.6474 - ETA: 18s - loss: 1.4873 - acc: 0.6487 - ETA: 18s - loss: 1.4831 - acc: 0.6497 - ETA: 18s - loss: 1.4801 - acc: 0.6503 - ETA: 18s - loss: 1.4741 - acc: 0.6515 - ETA: 18s - loss: 1.4704 - acc: 0.6522 - ETA: 18s - loss: 1.4662 - acc: 0.6532 - ETA: 17s - loss: 1.4624 - acc: 0.6539 - ETA: 17s - loss: 1.4597 - acc: 0.6544 - ETA: 17s - loss: 1.4536 - acc: 0.6559 - ETA: 17s - loss: 1.4501 - acc: 0.6563 - ETA: 17s - loss: 1.4471 - acc: 0.6570 - ETA: 17s - loss: 1.4438 - acc: 0.6577 - ETA: 17s - loss: 1.4392 - acc: 0.6586 - ETA: 16s - loss: 1.4369 - acc: 0.6588 - ETA: 16s - loss: 1.4309 - acc: 0.6604 - ETA: 16s - loss: 1.4289 - acc: 0.6611 - ETA: 16s - loss: 1.4269 - acc: 0.6615 - ETA: 16s - loss: 1.4218 - acc: 0.6626 - ETA: 16s - loss: 1.4193 - acc: 0.6630 - ETA: 16s - loss: 1.4198 - acc: 0.6625 - ETA: 16s - loss: 1.4151 - acc: 0.6631 - ETA: 15s - loss: 1.4109 - acc: 0.6642 - ETA: 15s - loss: 1.4065 - acc: 0.6655 - ETA: 15s - loss: 1.4040 - acc: 0.6661 - ETA: 15s - loss: 1.4011 - acc: 0.6661 - ETA: 15s - loss: 1.3969 - acc: 0.6667 - ETA: 15s - loss: 1.3950 - acc: 0.6673 - ETA: 15s - loss: 1.3939 - acc: 0.6674 - ETA: 14s - loss: 1.3906 - acc: 0.6682 - ETA: 14s - loss: 1.3875 - acc: 0.6688 - ETA: 14s - loss: 1.3841 - acc: 0.6696 - ETA: 14s - loss: 1.3800 - acc: 0.6706 - ETA: 14s - loss: 1.3761 - acc: 0.6714 - ETA: 14s - loss: 1.3770 - acc: 0.6711 - ETA: 14s - loss: 1.3758 - acc: 0.6714 - ETA: 14s - loss: 1.3737 - acc: 0.6718 - ETA: 13s - loss: 1.3698 - acc: 0.6723 - ETA: 13s - loss: 1.3667 - acc: 0.6726 - ETA: 13s - loss: 1.3634 - acc: 0.6734 - ETA: 13s - loss: 1.3593 - acc: 0.6742 - ETA: 13s - loss: 1.3553 - acc: 0.6745 - ETA: 13s - loss: 1.3522 - acc: 0.6748 - ETA: 13s - loss: 1.3510 - acc: 0.6751 - ETA: 13s - loss: 1.3478 - acc: 0.6754 - ETA: 13s - loss: 1.3449 - acc: 0.6755 - ETA: 13s - loss: 1.3428 - acc: 0.6760 - ETA: 13s - loss: 1.3414 - acc: 0.6761 - ETA: 12s - loss: 1.3376 - acc: 0.6768 - ETA: 12s - loss: 1.3353 - acc: 0.6773 - ETA: 12s - loss: 1.3350 - acc: 0.6772 - ETA: 12s - loss: 1.3318 - acc: 0.6777 - ETA: 12s - loss: 1.3279 - acc: 0.6784 - ETA: 12s - loss: 1.3256 - acc: 0.6789 - ETA: 11s - loss: 1.3218 - acc: 0.6798 - ETA: 11s - loss: 1.3203 - acc: 0.6799 - ETA: 11s - loss: 1.3168 - acc: 0.6806 - ETA: 11s - loss: 1.3143 - acc: 0.6812 - ETA: 11s - loss: 1.3123 - acc: 0.6817 - ETA: 11s - loss: 1.3075 - acc: 0.6829 - ETA: 11s - loss: 1.3066 - acc: 0.6832 - ETA: 10s - loss: 1.3043 - acc: 0.6837 - ETA: 10s - loss: 1.3021 - acc: 0.6843 - ETA: 10s - loss: 1.3002 - acc: 0.6851 - ETA: 10s - loss: 1.2983 - acc: 0.6856 - ETA: 10s - loss: 1.2982 - acc: 0.6852 - ETA: 10s - loss: 1.2976 - acc: 0.6851 - ETA: 9s - loss: 1.2970 - acc: 0.6856  - ETA: 9s - loss: 1.2955 - acc: 0.6858 - ETA: 9s - loss: 1.2938 - acc: 0.6860 - ETA: 9s - loss: 1.2938 - acc: 0.6859 - ETA: 9s - loss: 1.2908 - acc: 0.6863 - ETA: 9s - loss: 1.2867 - acc: 0.6873 - ETA: 9s - loss: 1.2834 - acc: 0.6879 - ETA: 8s - loss: 1.2797 - acc: 0.6885 - ETA: 8s - loss: 1.2783 - acc: 0.6887 - ETA: 8s - loss: 1.2752 - acc: 0.6893 - ETA: 8s - loss: 1.2732 - acc: 0.6899 - ETA: 8s - loss: 1.2700 - acc: 0.6907 - ETA: 8s - loss: 1.2678 - acc: 0.6913 - ETA: 8s - loss: 1.2646 - acc: 0.6918 - ETA: 7s - loss: 1.2631 - acc: 0.6921 - ETA: 7s - loss: 1.2597 - acc: 0.6927 - ETA: 7s - loss: 1.2578 - acc: 0.6930 - ETA: 7s - loss: 1.2546 - acc: 0.6936 - ETA: 6s - loss: 1.2534 - acc: 0.6936 - ETA: 6s - loss: 1.2509 - acc: 0.6944 - ETA: 6s - loss: 1.2474 - acc: 0.6951 - ETA: 6s - loss: 1.2471 - acc: 0.6948 - ETA: 6s - loss: 1.2449 - acc: 0.6951 - ETA: 6s - loss: 1.2435 - acc: 0.6957 - ETA: 6s - loss: 1.2415 - acc: 0.6960 - ETA: 6s - loss: 1.2388 - acc: 0.6966 - ETA: 5s - loss: 1.2375 - acc: 0.6967 - ETA: 5s - loss: 1.2359 - acc: 0.6971 - ETA: 5s - loss: 1.2375 - acc: 0.6964 - ETA: 5s - loss: 1.2347 - acc: 0.6969 - ETA: 5s - loss: 1.2332 - acc: 0.6975 - ETA: 5s - loss: 1.2332 - acc: 0.6978 - ETA: 5s - loss: 1.2324 - acc: 0.6980 - ETA: 4s - loss: 1.2301 - acc: 0.6985 - ETA: 4s - loss: 1.2287 - acc: 0.6985 - ETA: 4s - loss: 1.2259 - acc: 0.6992 - ETA: 4s - loss: 1.2227 - acc: 0.7000 - ETA: 4s - loss: 1.2210 - acc: 0.7002 - ETA: 4s - loss: 1.2212 - acc: 0.6998 - ETA: 4s - loss: 1.2206 - acc: 0.6997 - ETA: 3s - loss: 1.2207 - acc: 0.6998 - ETA: 3s - loss: 1.2188 - acc: 0.7003 - ETA: 3s - loss: 1.2158 - acc: 0.7011 - ETA: 3s - loss: 1.2146 - acc: 0.7010 - ETA: 3s - loss: 1.2135 - acc: 0.7008 - ETA: 3s - loss: 1.2110 - acc: 0.7015 - ETA: 3s - loss: 1.2098 - acc: 0.7018 - ETA: 2s - loss: 1.2076 - acc: 0.7022 - ETA: 2s - loss: 1.2058 - acc: 0.7026 - ETA: 2s - loss: 1.2041 - acc: 0.7027 - ETA: 2s - loss: 1.2033 - acc: 0.7027 - ETA: 2s - loss: 1.2015 - acc: 0.7028 - ETA: 2s - loss: 1.1993 - acc: 0.7032 - ETA: 2s - loss: 1.1977 - acc: 0.7031 - ETA: 2s - loss: 1.1970 - acc: 0.7033 - ETA: 1s - loss: 1.1958 - acc: 0.7036 - ETA: 1s - loss: 1.1947 - acc: 0.7037 - ETA: 1s - loss: 1.1925 - acc: 0.7042 - ETA: 1s - loss: 1.1915 - acc: 0.7043 - ETA: 1s - loss: 1.1895 - acc: 0.7045 - ETA: 1s - loss: 1.1886 - acc: 0.7049 - ETA: 1s - loss: 1.1861 - acc: 0.7054 - ETA: 0s - loss: 1.1834 - acc: 0.7060 - ETA: 0s - loss: 1.1806 - acc: 0.7066 - ETA: 0s - loss: 1.1800 - acc: 0.7067 - ETA: 0s - loss: 1.1775 - acc: 0.7071 - ETA: 0s - loss: 1.1760 - acc: 0.7073 - ETA: 0s - loss: 1.1744 - acc: 0.7077 - ETA: 0s - loss: 1.1731 - acc: 0.7077Epoch 00000: val_loss improved from inf to 0.64072, saving model to weights.best.InceptionV3.hdf5
    6680/6680 [==============================] - 45s - loss: 1.1717 - acc: 0.7075 - val_loss: 0.6407 - val_acc: 0.8096
    Epoch 2/20
    6620/6680 [============================>.] - ETA: 12s - loss: 0.5847 - acc: 0.9000 - ETA: 6s - loss: 0.4102 - acc: 0.8900  - ETA: 6s - loss: 0.4246 - acc: 0.8875 - ETA: 6s - loss: 0.4827 - acc: 0.8682 - ETA: 6s - loss: 0.4443 - acc: 0.8714 - ETA: 5s - loss: 0.4014 - acc: 0.8853 - ETA: 5s - loss: 0.3667 - acc: 0.8975 - ETA: 5s - loss: 0.3532 - acc: 0.9000 - ETA: 5s - loss: 0.3571 - acc: 0.9000 - ETA: 5s - loss: 0.3674 - acc: 0.8948 - ETA: 5s - loss: 0.3645 - acc: 0.8984 - ETA: 5s - loss: 0.3623 - acc: 0.8986 - ETA: 5s - loss: 0.3797 - acc: 0.8908 - ETA: 5s - loss: 0.3864 - acc: 0.8902 - ETA: 5s - loss: 0.3717 - acc: 0.8966 - ETA: 5s - loss: 0.3750 - acc: 0.8936 - ETA: 5s - loss: 0.3792 - acc: 0.8930 - ETA: 5s - loss: 0.3805 - acc: 0.8906 - ETA: 4s - loss: 0.3977 - acc: 0.8833 - ETA: 4s - loss: 0.4269 - acc: 0.8779 - ETA: 4s - loss: 0.4385 - acc: 0.8754 - ETA: 4s - loss: 0.4383 - acc: 0.8750 - ETA: 4s - loss: 0.4352 - acc: 0.8754 - ETA: 4s - loss: 0.4303 - acc: 0.8767 - ETA: 4s - loss: 0.4381 - acc: 0.8731 - ETA: 4s - loss: 0.4491 - acc: 0.8710 - ETA: 4s - loss: 0.4538 - acc: 0.8702 - ETA: 4s - loss: 0.4567 - acc: 0.8678 - ETA: 4s - loss: 0.4571 - acc: 0.8661 - ETA: 4s - loss: 0.4598 - acc: 0.8634 - ETA: 4s - loss: 0.4624 - acc: 0.8635 - ETA: 4s - loss: 0.4722 - acc: 0.8596 - ETA: 4s - loss: 0.4679 - acc: 0.8617 - ETA: 3s - loss: 0.4620 - acc: 0.8642 - ETA: 3s - loss: 0.4602 - acc: 0.8642 - ETA: 3s - loss: 0.4662 - acc: 0.8612 - ETA: 3s - loss: 0.4624 - acc: 0.8621 - ETA: 3s - loss: 0.4693 - acc: 0.8596 - ETA: 3s - loss: 0.4731 - acc: 0.8577 - ETA: 3s - loss: 0.4724 - acc: 0.8575 - ETA: 3s - loss: 0.4742 - acc: 0.8562 - ETA: 3s - loss: 0.4739 - acc: 0.8560 - ETA: 3s - loss: 0.4761 - acc: 0.8544 - ETA: 3s - loss: 0.4759 - acc: 0.8543 - ETA: 3s - loss: 0.4794 - acc: 0.8542 - ETA: 3s - loss: 0.4775 - acc: 0.8554 - ETA: 3s - loss: 0.4760 - acc: 0.8553 - ETA: 3s - loss: 0.4793 - acc: 0.8551 - ETA: 2s - loss: 0.4751 - acc: 0.8559 - ETA: 2s - loss: 0.4750 - acc: 0.8564 - ETA: 2s - loss: 0.4781 - acc: 0.8557 - ETA: 2s - loss: 0.4739 - acc: 0.8558 - ETA: 2s - loss: 0.4752 - acc: 0.8554 - ETA: 2s - loss: 0.4759 - acc: 0.8550 - ETA: 2s - loss: 0.4809 - acc: 0.8546 - ETA: 2s - loss: 0.4773 - acc: 0.8559 - ETA: 2s - loss: 0.4804 - acc: 0.8544 - ETA: 2s - loss: 0.4764 - acc: 0.8551 - ETA: 2s - loss: 0.4781 - acc: 0.8547 - ETA: 2s - loss: 0.4783 - acc: 0.8547 - ETA: 2s - loss: 0.4801 - acc: 0.8546 - ETA: 2s - loss: 0.4777 - acc: 0.8557 - ETA: 1s - loss: 0.4748 - acc: 0.8560 - ETA: 1s - loss: 0.4730 - acc: 0.8573 - ETA: 1s - loss: 0.4733 - acc: 0.8574 - ETA: 1s - loss: 0.4751 - acc: 0.8573 - ETA: 1s - loss: 0.4715 - acc: 0.8576 - ETA: 1s - loss: 0.4743 - acc: 0.8568 - ETA: 1s - loss: 0.4736 - acc: 0.8563 - ETA: 1s - loss: 0.4778 - acc: 0.8553 - ETA: 1s - loss: 0.4767 - acc: 0.8555 - ETA: 1s - loss: 0.4781 - acc: 0.8552 - ETA: 1s - loss: 0.4755 - acc: 0.8555 - ETA: 1s - loss: 0.4768 - acc: 0.8550 - ETA: 1s - loss: 0.4748 - acc: 0.8559 - ETA: 1s - loss: 0.4756 - acc: 0.8560 - ETA: 1s - loss: 0.4767 - acc: 0.8558 - ETA: 1s - loss: 0.4751 - acc: 0.8572 - ETA: 0s - loss: 0.4747 - acc: 0.8572 - ETA: 0s - loss: 0.4719 - acc: 0.8575 - ETA: 0s - loss: 0.4732 - acc: 0.8577 - ETA: 0s - loss: 0.4722 - acc: 0.8583 - ETA: 0s - loss: 0.4694 - acc: 0.8589 - ETA: 0s - loss: 0.4710 - acc: 0.8583 - ETA: 0s - loss: 0.4690 - acc: 0.8580 - ETA: 0s - loss: 0.4688 - acc: 0.8577 - ETA: 0s - loss: 0.4701 - acc: 0.8581 - ETA: 0s - loss: 0.4707 - acc: 0.8579 - ETA: 0s - loss: 0.4709 - acc: 0.8579 - ETA: 0s - loss: 0.4701 - acc: 0.8575 - ETA: 0s - loss: 0.4717 - acc: 0.8570 - ETA: 0s - loss: 0.4724 - acc: 0.8563 - ETA: 0s - loss: 0.4723 - acc: 0.8562Epoch 00001: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.4719 - acc: 0.8561 - val_loss: 0.6837 - val_acc: 0.8108
    Epoch 3/20
    6600/6680 [============================>.] - ETA: 6s - loss: 0.4983 - acc: 0.9500 - ETA: 5s - loss: 0.2428 - acc: 0.9375 - ETA: 5s - loss: 0.2301 - acc: 0.9437 - ETA: 5s - loss: 0.2256 - acc: 0.9458 - ETA: 5s - loss: 0.2608 - acc: 0.9312 - ETA: 4s - loss: 0.2755 - acc: 0.9275 - ETA: 4s - loss: 0.2882 - acc: 0.9208 - ETA: 4s - loss: 0.2736 - acc: 0.9232 - ETA: 4s - loss: 0.2701 - acc: 0.9210 - ETA: 4s - loss: 0.2706 - acc: 0.9200 - ETA: 4s - loss: 0.2974 - acc: 0.9115 - ETA: 4s - loss: 0.2897 - acc: 0.9140 - ETA: 4s - loss: 0.2809 - acc: 0.9138 - ETA: 4s - loss: 0.2858 - acc: 0.9147 - ETA: 4s - loss: 0.2916 - acc: 0.9120 - ETA: 4s - loss: 0.3144 - acc: 0.9043 - ETA: 4s - loss: 0.3110 - acc: 0.9056 - ETA: 4s - loss: 0.3097 - acc: 0.9045 - ETA: 4s - loss: 0.3139 - acc: 0.9022 - ETA: 4s - loss: 0.3054 - acc: 0.9055 - ETA: 3s - loss: 0.3198 - acc: 0.9006 - ETA: 3s - loss: 0.3318 - acc: 0.8994 - ETA: 3s - loss: 0.3286 - acc: 0.9000 - ETA: 3s - loss: 0.3272 - acc: 0.9011 - ETA: 3s - loss: 0.3312 - acc: 0.8978 - ETA: 3s - loss: 0.3307 - acc: 0.8964 - ETA: 3s - loss: 0.3267 - acc: 0.8965 - ETA: 3s - loss: 0.3226 - acc: 0.8981 - ETA: 3s - loss: 0.3280 - acc: 0.8959 - ETA: 3s - loss: 0.3246 - acc: 0.8978 - ETA: 3s - loss: 0.3310 - acc: 0.8966 - ETA: 3s - loss: 0.3339 - acc: 0.8950 - ETA: 3s - loss: 0.3328 - acc: 0.8952 - ETA: 3s - loss: 0.3461 - acc: 0.8918 - ETA: 3s - loss: 0.3441 - acc: 0.8920 - ETA: 3s - loss: 0.3458 - acc: 0.8915 - ETA: 3s - loss: 0.3430 - acc: 0.8904 - ETA: 2s - loss: 0.3446 - acc: 0.8903 - ETA: 2s - loss: 0.3441 - acc: 0.8905 - ETA: 2s - loss: 0.3420 - acc: 0.8908 - ETA: 2s - loss: 0.3386 - acc: 0.8913 - ETA: 2s - loss: 0.3368 - acc: 0.8919 - ETA: 2s - loss: 0.3356 - acc: 0.8927 - ETA: 2s - loss: 0.3366 - acc: 0.8929 - ETA: 2s - loss: 0.3353 - acc: 0.8933 - ETA: 2s - loss: 0.3402 - acc: 0.8929 - ETA: 2s - loss: 0.3405 - acc: 0.8925 - ETA: 2s - loss: 0.3411 - acc: 0.8924 - ETA: 2s - loss: 0.3413 - acc: 0.8926 - ETA: 2s - loss: 0.3429 - acc: 0.8924 - ETA: 2s - loss: 0.3520 - acc: 0.8911 - ETA: 2s - loss: 0.3514 - acc: 0.8912 - ETA: 2s - loss: 0.3500 - acc: 0.8909 - ETA: 1s - loss: 0.3508 - acc: 0.8909 - ETA: 1s - loss: 0.3523 - acc: 0.8910 - ETA: 1s - loss: 0.3509 - acc: 0.8912 - ETA: 1s - loss: 0.3535 - acc: 0.8909 - ETA: 1s - loss: 0.3520 - acc: 0.8915 - ETA: 1s - loss: 0.3510 - acc: 0.8914 - ETA: 1s - loss: 0.3504 - acc: 0.8918 - ETA: 1s - loss: 0.3507 - acc: 0.8917 - ETA: 1s - loss: 0.3531 - acc: 0.8908 - ETA: 1s - loss: 0.3582 - acc: 0.8893 - ETA: 1s - loss: 0.3570 - acc: 0.8895 - ETA: 1s - loss: 0.3569 - acc: 0.8902 - ETA: 1s - loss: 0.3543 - acc: 0.8909 - ETA: 1s - loss: 0.3558 - acc: 0.8903 - ETA: 1s - loss: 0.3556 - acc: 0.8902 - ETA: 1s - loss: 0.3554 - acc: 0.8896 - ETA: 1s - loss: 0.3571 - acc: 0.8894 - ETA: 0s - loss: 0.3572 - acc: 0.8897 - ETA: 0s - loss: 0.3571 - acc: 0.8899 - ETA: 0s - loss: 0.3610 - acc: 0.8891 - ETA: 0s - loss: 0.3579 - acc: 0.8901 - ETA: 0s - loss: 0.3567 - acc: 0.8899 - ETA: 0s - loss: 0.3575 - acc: 0.8899 - ETA: 0s - loss: 0.3560 - acc: 0.8905 - ETA: 0s - loss: 0.3565 - acc: 0.8903 - ETA: 0s - loss: 0.3560 - acc: 0.8905 - ETA: 0s - loss: 0.3551 - acc: 0.8904 - ETA: 0s - loss: 0.3541 - acc: 0.8904 - ETA: 0s - loss: 0.3534 - acc: 0.8903 - ETA: 0s - loss: 0.3546 - acc: 0.8899 - ETA: 0s - loss: 0.3551 - acc: 0.8899 - ETA: 0s - loss: 0.3541 - acc: 0.8903 - ETA: 0s - loss: 0.3546 - acc: 0.8902Epoch 00002: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.3582 - acc: 0.8897 - val_loss: 0.6450 - val_acc: 0.8383
    Epoch 4/20
    6660/6680 [============================>.] - ETA: 5s - loss: 0.2810 - acc: 0.9000 - ETA: 4s - loss: 0.1850 - acc: 0.9300 - ETA: 4s - loss: 0.1877 - acc: 0.9278 - ETA: 4s - loss: 0.2624 - acc: 0.9154 - ETA: 4s - loss: 0.3056 - acc: 0.9059 - ETA: 4s - loss: 0.2990 - acc: 0.9000 - ETA: 4s - loss: 0.2731 - acc: 0.9062 - ETA: 4s - loss: 0.2535 - acc: 0.9111 - ETA: 4s - loss: 0.2470 - acc: 0.9145 - ETA: 4s - loss: 0.2451 - acc: 0.9186 - ETA: 4s - loss: 0.2346 - acc: 0.9231 - ETA: 4s - loss: 0.2483 - acc: 0.9186 - ETA: 4s - loss: 0.2639 - acc: 0.9196 - ETA: 4s - loss: 0.2633 - acc: 0.9194 - ETA: 4s - loss: 0.2562 - acc: 0.9202 - ETA: 4s - loss: 0.2472 - acc: 0.9214 - ETA: 4s - loss: 0.2460 - acc: 0.9217 - ETA: 4s - loss: 0.2552 - acc: 0.9195 - ETA: 4s - loss: 0.2555 - acc: 0.9206 - ETA: 4s - loss: 0.2525 - acc: 0.9225 - ETA: 4s - loss: 0.2485 - acc: 0.9227 - ETA: 4s - loss: 0.2402 - acc: 0.9247 - ETA: 3s - loss: 0.2487 - acc: 0.9232 - ETA: 3s - loss: 0.2456 - acc: 0.9244 - ETA: 3s - loss: 0.2450 - acc: 0.9247 - ETA: 3s - loss: 0.2549 - acc: 0.9228 - ETA: 3s - loss: 0.2514 - acc: 0.9237 - ETA: 3s - loss: 0.2598 - acc: 0.9219 - ETA: 3s - loss: 0.2587 - acc: 0.9218 - ETA: 3s - loss: 0.2541 - acc: 0.9231 - ETA: 3s - loss: 0.2563 - acc: 0.9229 - ETA: 3s - loss: 0.2587 - acc: 0.9218 - ETA: 3s - loss: 0.2623 - acc: 0.9190 - ETA: 3s - loss: 0.2624 - acc: 0.9181 - ETA: 3s - loss: 0.2629 - acc: 0.9181 - ETA: 3s - loss: 0.2649 - acc: 0.9184 - ETA: 3s - loss: 0.2653 - acc: 0.9192 - ETA: 3s - loss: 0.2658 - acc: 0.9191 - ETA: 3s - loss: 0.2648 - acc: 0.9183 - ETA: 3s - loss: 0.2658 - acc: 0.9183 - ETA: 3s - loss: 0.2666 - acc: 0.9182 - ETA: 3s - loss: 0.2673 - acc: 0.9182 - ETA: 3s - loss: 0.2649 - acc: 0.9192 - ETA: 3s - loss: 0.2658 - acc: 0.9185 - ETA: 3s - loss: 0.2657 - acc: 0.9185 - ETA: 3s - loss: 0.2651 - acc: 0.9178 - ETA: 3s - loss: 0.2663 - acc: 0.9174 - ETA: 2s - loss: 0.2651 - acc: 0.9174 - ETA: 2s - loss: 0.2654 - acc: 0.9177 - ETA: 2s - loss: 0.2648 - acc: 0.9180 - ETA: 2s - loss: 0.2648 - acc: 0.9177 - ETA: 2s - loss: 0.2667 - acc: 0.9171 - ETA: 2s - loss: 0.2659 - acc: 0.9171 - ETA: 2s - loss: 0.2650 - acc: 0.9179 - ETA: 2s - loss: 0.2635 - acc: 0.9182 - ETA: 2s - loss: 0.2619 - acc: 0.9187 - ETA: 2s - loss: 0.2622 - acc: 0.9184 - ETA: 2s - loss: 0.2627 - acc: 0.9184 - ETA: 2s - loss: 0.2632 - acc: 0.9186 - ETA: 2s - loss: 0.2622 - acc: 0.9191 - ETA: 2s - loss: 0.2597 - acc: 0.9198 - ETA: 2s - loss: 0.2607 - acc: 0.9195 - ETA: 2s - loss: 0.2668 - acc: 0.9190 - ETA: 2s - loss: 0.2709 - acc: 0.9182 - ETA: 2s - loss: 0.2711 - acc: 0.9165 - ETA: 2s - loss: 0.2715 - acc: 0.9158 - ETA: 2s - loss: 0.2714 - acc: 0.9160 - ETA: 1s - loss: 0.2702 - acc: 0.9159 - ETA: 1s - loss: 0.2722 - acc: 0.9156 - ETA: 1s - loss: 0.2700 - acc: 0.9162 - ETA: 1s - loss: 0.2705 - acc: 0.9157 - ETA: 1s - loss: 0.2701 - acc: 0.9155 - ETA: 1s - loss: 0.2686 - acc: 0.9156 - ETA: 1s - loss: 0.2691 - acc: 0.9154 - ETA: 1s - loss: 0.2711 - acc: 0.9149 - ETA: 1s - loss: 0.2747 - acc: 0.9147 - ETA: 1s - loss: 0.2795 - acc: 0.9145 - ETA: 1s - loss: 0.2788 - acc: 0.9147 - ETA: 1s - loss: 0.2786 - acc: 0.9146 - ETA: 1s - loss: 0.2774 - acc: 0.9152 - ETA: 1s - loss: 0.2778 - acc: 0.9150 - ETA: 1s - loss: 0.2784 - acc: 0.9150 - ETA: 0s - loss: 0.2774 - acc: 0.9156 - ETA: 0s - loss: 0.2787 - acc: 0.9154 - ETA: 0s - loss: 0.2773 - acc: 0.9154 - ETA: 0s - loss: 0.2766 - acc: 0.9156 - ETA: 0s - loss: 0.2775 - acc: 0.9151 - ETA: 0s - loss: 0.2753 - acc: 0.9158 - ETA: 0s - loss: 0.2742 - acc: 0.9162 - ETA: 0s - loss: 0.2750 - acc: 0.9158 - ETA: 0s - loss: 0.2782 - acc: 0.9153 - ETA: 0s - loss: 0.2772 - acc: 0.9155 - ETA: 0s - loss: 0.2762 - acc: 0.9157 - ETA: 0s - loss: 0.2774 - acc: 0.9152 - ETA: 0s - loss: 0.2775 - acc: 0.9149 - ETA: 0s - loss: 0.2816 - acc: 0.9141 - ETA: 0s - loss: 0.2838 - acc: 0.9132 - ETA: 0s - loss: 0.2840 - acc: 0.9129 - ETA: 0s - loss: 0.2876 - acc: 0.9122 - ETA: 0s - loss: 0.2873 - acc: 0.9124 - ETA: 0s - loss: 0.2871 - acc: 0.9123 - ETA: 0s - loss: 0.2877 - acc: 0.9122Epoch 00003: val_loss did not improve
    6680/6680 [==============================] - 6s - loss: 0.2870 - acc: 0.9124 - val_loss: 0.6777 - val_acc: 0.8491
    Epoch 5/20
    6660/6680 [============================>.] - ETA: 5s - loss: 0.1769 - acc: 0.9000 - ETA: 5s - loss: 0.2604 - acc: 0.9375 - ETA: 5s - loss: 0.2189 - acc: 0.9286 - ETA: 5s - loss: 0.2409 - acc: 0.9100 - ETA: 5s - loss: 0.2046 - acc: 0.9269 - ETA: 5s - loss: 0.2061 - acc: 0.9281 - ETA: 5s - loss: 0.2487 - acc: 0.9158 - ETA: 5s - loss: 0.2563 - acc: 0.9205 - ETA: 5s - loss: 0.2385 - acc: 0.9260 - ETA: 5s - loss: 0.2422 - acc: 0.9268 - ETA: 5s - loss: 0.2265 - acc: 0.9306 - ETA: 5s - loss: 0.2375 - acc: 0.9265 - ETA: 5s - loss: 0.2225 - acc: 0.9324 - ETA: 5s - loss: 0.2307 - acc: 0.9300 - ETA: 5s - loss: 0.2228 - acc: 0.9326 - ETA: 5s - loss: 0.2460 - acc: 0.9315 - ETA: 5s - loss: 0.2466 - acc: 0.9296 - ETA: 5s - loss: 0.2388 - acc: 0.9308 - ETA: 4s - loss: 0.2331 - acc: 0.9312 - ETA: 4s - loss: 0.2353 - acc: 0.9322 - ETA: 4s - loss: 0.2232 - acc: 0.9365 - ETA: 4s - loss: 0.2210 - acc: 0.9348 - ETA: 4s - loss: 0.2202 - acc: 0.9350 - ETA: 4s - loss: 0.2234 - acc: 0.9329 - ETA: 4s - loss: 0.2246 - acc: 0.9329 - ETA: 4s - loss: 0.2253 - acc: 0.9331 - ETA: 4s - loss: 0.2266 - acc: 0.9339 - ETA: 4s - loss: 0.2204 - acc: 0.9347 - ETA: 4s - loss: 0.2192 - acc: 0.9341 - ETA: 4s - loss: 0.2182 - acc: 0.9330 - ETA: 4s - loss: 0.2176 - acc: 0.9335 - ETA: 4s - loss: 0.2134 - acc: 0.9347 - ETA: 3s - loss: 0.2241 - acc: 0.9295 - ETA: 3s - loss: 0.2229 - acc: 0.9298 - ETA: 3s - loss: 0.2214 - acc: 0.9299 - ETA: 3s - loss: 0.2175 - acc: 0.9309 - ETA: 3s - loss: 0.2280 - acc: 0.9292 - ETA: 3s - loss: 0.2309 - acc: 0.9285 - ETA: 3s - loss: 0.2341 - acc: 0.9274 - ETA: 3s - loss: 0.2335 - acc: 0.9268 - ETA: 3s - loss: 0.2338 - acc: 0.9262 - ETA: 3s - loss: 0.2331 - acc: 0.9256 - ETA: 3s - loss: 0.2312 - acc: 0.9261 - ETA: 3s - loss: 0.2290 - acc: 0.9266 - ETA: 3s - loss: 0.2271 - acc: 0.9268 - ETA: 3s - loss: 0.2283 - acc: 0.9262 - ETA: 3s - loss: 0.2275 - acc: 0.9260 - ETA: 3s - loss: 0.2302 - acc: 0.9248 - ETA: 3s - loss: 0.2311 - acc: 0.9250 - ETA: 3s - loss: 0.2304 - acc: 0.9248 - ETA: 3s - loss: 0.2318 - acc: 0.9253 - ETA: 2s - loss: 0.2294 - acc: 0.9258 - ETA: 2s - loss: 0.2268 - acc: 0.9268 - ETA: 2s - loss: 0.2259 - acc: 0.9269 - ETA: 2s - loss: 0.2268 - acc: 0.9265 - ETA: 2s - loss: 0.2272 - acc: 0.9266 - ETA: 2s - loss: 0.2264 - acc: 0.9270 - ETA: 2s - loss: 0.2284 - acc: 0.9271 - ETA: 2s - loss: 0.2303 - acc: 0.9266 - ETA: 2s - loss: 0.2310 - acc: 0.9265 - ETA: 2s - loss: 0.2320 - acc: 0.9266 - ETA: 2s - loss: 0.2307 - acc: 0.9267 - ETA: 2s - loss: 0.2299 - acc: 0.9265 - ETA: 2s - loss: 0.2287 - acc: 0.9269 - ETA: 2s - loss: 0.2279 - acc: 0.9272 - ETA: 2s - loss: 0.2278 - acc: 0.9273 - ETA: 2s - loss: 0.2268 - acc: 0.9274 - ETA: 2s - loss: 0.2262 - acc: 0.9275 - ETA: 2s - loss: 0.2263 - acc: 0.9276 - ETA: 2s - loss: 0.2270 - acc: 0.9276 - ETA: 1s - loss: 0.2283 - acc: 0.9273 - ETA: 1s - loss: 0.2296 - acc: 0.9271 - ETA: 1s - loss: 0.2295 - acc: 0.9274 - ETA: 1s - loss: 0.2309 - acc: 0.9269 - ETA: 1s - loss: 0.2317 - acc: 0.9274 - ETA: 1s - loss: 0.2296 - acc: 0.9281 - ETA: 1s - loss: 0.2285 - acc: 0.9284 - ETA: 1s - loss: 0.2301 - acc: 0.9282 - ETA: 1s - loss: 0.2290 - acc: 0.9283 - ETA: 1s - loss: 0.2298 - acc: 0.9275 - ETA: 1s - loss: 0.2333 - acc: 0.9262 - ETA: 1s - loss: 0.2336 - acc: 0.9261 - ETA: 1s - loss: 0.2336 - acc: 0.9260 - ETA: 1s - loss: 0.2333 - acc: 0.9261 - ETA: 1s - loss: 0.2358 - acc: 0.9255 - ETA: 1s - loss: 0.2393 - acc: 0.9244 - ETA: 1s - loss: 0.2392 - acc: 0.9243 - ETA: 1s - loss: 0.2392 - acc: 0.9240 - ETA: 0s - loss: 0.2399 - acc: 0.9238 - ETA: 0s - loss: 0.2391 - acc: 0.9240 - ETA: 0s - loss: 0.2376 - acc: 0.9247 - ETA: 0s - loss: 0.2417 - acc: 0.9240 - ETA: 0s - loss: 0.2405 - acc: 0.9242 - ETA: 0s - loss: 0.2425 - acc: 0.9241 - ETA: 0s - loss: 0.2453 - acc: 0.9234 - ETA: 0s - loss: 0.2463 - acc: 0.9234 - ETA: 0s - loss: 0.2464 - acc: 0.9236 - ETA: 0s - loss: 0.2445 - acc: 0.9243 - ETA: 0s - loss: 0.2442 - acc: 0.9246 - ETA: 0s - loss: 0.2431 - acc: 0.9249 - ETA: 0s - loss: 0.2432 - acc: 0.9249 - ETA: 0s - loss: 0.2415 - acc: 0.9255 - ETA: 0s - loss: 0.2392 - acc: 0.9263Epoch 00004: val_loss did not improve
    6680/6680 [==============================] - 6s - loss: 0.2398 - acc: 0.9263 - val_loss: 0.7026 - val_acc: 0.8467
    Epoch 6/20
    6640/6680 [============================>.] - ETA: 5s - loss: 0.0230 - acc: 1.0000 - ETA: 5s - loss: 0.0990 - acc: 0.9400 - ETA: 5s - loss: 0.1093 - acc: 0.9437 - ETA: 5s - loss: 0.1225 - acc: 0.9455 - ETA: 5s - loss: 0.1280 - acc: 0.9429 - ETA: 5s - loss: 0.1254 - acc: 0.9412 - ETA: 5s - loss: 0.1309 - acc: 0.9400 - ETA: 5s - loss: 0.1462 - acc: 0.9391 - ETA: 5s - loss: 0.1431 - acc: 0.9385 - ETA: 5s - loss: 0.1383 - acc: 0.9397 - ETA: 5s - loss: 0.1380 - acc: 0.9391 - ETA: 5s - loss: 0.1621 - acc: 0.9371 - ETA: 5s - loss: 0.1573 - acc: 0.9382 - ETA: 4s - loss: 0.1738 - acc: 0.9341 - ETA: 4s - loss: 0.1738 - acc: 0.9364 - ETA: 4s - loss: 0.1660 - acc: 0.9406 - ETA: 4s - loss: 0.1631 - acc: 0.9402 - ETA: 4s - loss: 0.1617 - acc: 0.9407 - ETA: 4s - loss: 0.1675 - acc: 0.9421 - ETA: 4s - loss: 0.1682 - acc: 0.9400 - ETA: 4s - loss: 0.1667 - acc: 0.9405 - ETA: 4s - loss: 0.1650 - acc: 0.9409 - ETA: 4s - loss: 0.1616 - acc: 0.9420 - ETA: 4s - loss: 0.1639 - acc: 0.9417 - ETA: 4s - loss: 0.1603 - acc: 0.9427 - ETA: 4s - loss: 0.1660 - acc: 0.9423 - ETA: 4s - loss: 0.1677 - acc: 0.9426 - ETA: 4s - loss: 0.1689 - acc: 0.9429 - ETA: 4s - loss: 0.1720 - acc: 0.9414 - ETA: 4s - loss: 0.1775 - acc: 0.9400 - ETA: 4s - loss: 0.1774 - acc: 0.9398 - ETA: 4s - loss: 0.1746 - acc: 0.9401 - ETA: 3s - loss: 0.1768 - acc: 0.9394 - ETA: 3s - loss: 0.1734 - acc: 0.9407 - ETA: 3s - loss: 0.1718 - acc: 0.9410 - ETA: 3s - loss: 0.1696 - acc: 0.9413 - ETA: 3s - loss: 0.1675 - acc: 0.9419 - ETA: 3s - loss: 0.1678 - acc: 0.9416 - ETA: 3s - loss: 0.1687 - acc: 0.9412 - ETA: 3s - loss: 0.1695 - acc: 0.9405 - ETA: 4s - loss: 0.1709 - acc: 0.9394 - ETA: 4s - loss: 0.1719 - acc: 0.9387 - ETA: 4s - loss: 0.1708 - acc: 0.9393 - ETA: 4s - loss: 0.1696 - acc: 0.9394 - ETA: 4s - loss: 0.1720 - acc: 0.9396 - ETA: 4s - loss: 0.1708 - acc: 0.9402 - ETA: 3s - loss: 0.1766 - acc: 0.9393 - ETA: 3s - loss: 0.1779 - acc: 0.9384 - ETA: 3s - loss: 0.1799 - acc: 0.9380 - ETA: 3s - loss: 0.1815 - acc: 0.9375 - ETA: 3s - loss: 0.1824 - acc: 0.9378 - ETA: 3s - loss: 0.1817 - acc: 0.9380 - ETA: 3s - loss: 0.1801 - acc: 0.9389 - ETA: 3s - loss: 0.1779 - acc: 0.9395 - ETA: 3s - loss: 0.1770 - acc: 0.9404 - ETA: 3s - loss: 0.1766 - acc: 0.9412 - ETA: 3s - loss: 0.1775 - acc: 0.9408 - ETA: 3s - loss: 0.1780 - acc: 0.9410 - ETA: 3s - loss: 0.1798 - acc: 0.9408 - ETA: 3s - loss: 0.1799 - acc: 0.9404 - ETA: 3s - loss: 0.1823 - acc: 0.9394 - ETA: 2s - loss: 0.1835 - acc: 0.9390 - ETA: 2s - loss: 0.1843 - acc: 0.9390 - ETA: 2s - loss: 0.1857 - acc: 0.9383 - ETA: 2s - loss: 0.1847 - acc: 0.9388 - ETA: 2s - loss: 0.1860 - acc: 0.9382 - ETA: 2s - loss: 0.1887 - acc: 0.9370 - ETA: 2s - loss: 0.1892 - acc: 0.9375 - ETA: 2s - loss: 0.1875 - acc: 0.9379 - ETA: 2s - loss: 0.1852 - acc: 0.9389 - ETA: 2s - loss: 0.1841 - acc: 0.9393 - ETA: 2s - loss: 0.1859 - acc: 0.9392 - ETA: 2s - loss: 0.1886 - acc: 0.9386 - ETA: 2s - loss: 0.1882 - acc: 0.9386 - ETA: 2s - loss: 0.1880 - acc: 0.9387 - ETA: 2s - loss: 0.1884 - acc: 0.9384 - ETA: 2s - loss: 0.1899 - acc: 0.9386 - ETA: 2s - loss: 0.1910 - acc: 0.9383 - ETA: 1s - loss: 0.1898 - acc: 0.9389 - ETA: 1s - loss: 0.1881 - acc: 0.9394 - ETA: 1s - loss: 0.1896 - acc: 0.9387 - ETA: 1s - loss: 0.1877 - acc: 0.9393 - ETA: 1s - loss: 0.1883 - acc: 0.9388 - ETA: 1s - loss: 0.1933 - acc: 0.9383 - ETA: 1s - loss: 0.1936 - acc: 0.9381 - ETA: 1s - loss: 0.1944 - acc: 0.9378 - ETA: 1s - loss: 0.1941 - acc: 0.9381 - ETA: 1s - loss: 0.1934 - acc: 0.9383 - ETA: 1s - loss: 0.1934 - acc: 0.9383 - ETA: 1s - loss: 0.1919 - acc: 0.9387 - ETA: 1s - loss: 0.1920 - acc: 0.9387 - ETA: 1s - loss: 0.1930 - acc: 0.9386 - ETA: 1s - loss: 0.1913 - acc: 0.9393 - ETA: 1s - loss: 0.1906 - acc: 0.9394 - ETA: 0s - loss: 0.1951 - acc: 0.9385 - ETA: 0s - loss: 0.1945 - acc: 0.9383 - ETA: 0s - loss: 0.1953 - acc: 0.9379 - ETA: 0s - loss: 0.1953 - acc: 0.9382 - ETA: 0s - loss: 0.1931 - acc: 0.9391 - ETA: 0s - loss: 0.1939 - acc: 0.9387 - ETA: 0s - loss: 0.1950 - acc: 0.9387 - ETA: 0s - loss: 0.1953 - acc: 0.9385 - ETA: 0s - loss: 0.1974 - acc: 0.9378 - ETA: 0s - loss: 0.1997 - acc: 0.9371 - ETA: 0s - loss: 0.1995 - acc: 0.9373 - ETA: 0s - loss: 0.2002 - acc: 0.9371 - ETA: 0s - loss: 0.1996 - acc: 0.9373 - ETA: 0s - loss: 0.1986 - acc: 0.9374 - ETA: 0s - loss: 0.1988 - acc: 0.9372Epoch 00005: val_loss did not improve
    6680/6680 [==============================] - 6s - loss: 0.1991 - acc: 0.9370 - val_loss: 0.7065 - val_acc: 0.8479
    Epoch 7/20
    6660/6680 [============================>.] - ETA: 4s - loss: 0.2414 - acc: 0.8500 - ETA: 4s - loss: 0.2107 - acc: 0.9200 - ETA: 4s - loss: 0.1765 - acc: 0.9222 - ETA: 4s - loss: 0.1447 - acc: 0.9346 - ETA: 4s - loss: 0.1574 - acc: 0.9382 - ETA: 4s - loss: 0.1572 - acc: 0.9405 - ETA: 4s - loss: 0.1473 - acc: 0.9420 - ETA: 4s - loss: 0.1400 - acc: 0.9466 - ETA: 4s - loss: 0.1562 - acc: 0.9470 - ETA: 4s - loss: 0.1472 - acc: 0.9500 - ETA: 4s - loss: 0.1413 - acc: 0.9512 - ETA: 4s - loss: 0.1521 - acc: 0.9500 - ETA: 4s - loss: 0.1506 - acc: 0.9490 - ETA: 4s - loss: 0.1494 - acc: 0.9491 - ETA: 4s - loss: 0.1579 - acc: 0.9474 - ETA: 4s - loss: 0.1631 - acc: 0.9459 - ETA: 4s - loss: 0.1645 - acc: 0.9423 - ETA: 3s - loss: 0.1651 - acc: 0.9428 - ETA: 3s - loss: 0.1671 - acc: 0.9432 - ETA: 3s - loss: 0.1626 - acc: 0.9442 - ETA: 3s - loss: 0.1585 - acc: 0.9457 - ETA: 3s - loss: 0.1529 - acc: 0.9476 - ETA: 3s - loss: 0.1488 - acc: 0.9483 - ETA: 3s - loss: 0.1554 - acc: 0.9462 - ETA: 3s - loss: 0.1526 - acc: 0.9469 - ETA: 3s - loss: 0.1505 - acc: 0.9480 - ETA: 3s - loss: 0.1570 - acc: 0.9481 - ETA: 3s - loss: 0.1559 - acc: 0.9486 - ETA: 3s - loss: 0.1534 - acc: 0.9496 - ETA: 3s - loss: 0.1500 - acc: 0.9504 - ETA: 3s - loss: 0.1498 - acc: 0.9504 - ETA: 3s - loss: 0.1479 - acc: 0.9508 - ETA: 3s - loss: 0.1450 - acc: 0.9519 - ETA: 2s - loss: 0.1505 - acc: 0.9500 - ETA: 2s - loss: 0.1552 - acc: 0.9485 - ETA: 2s - loss: 0.1542 - acc: 0.9486 - ETA: 2s - loss: 0.1590 - acc: 0.9469 - ETA: 2s - loss: 0.1572 - acc: 0.9473 - ETA: 2s - loss: 0.1559 - acc: 0.9474 - ETA: 2s - loss: 0.1574 - acc: 0.9465 - ETA: 2s - loss: 0.1547 - acc: 0.9475 - ETA: 2s - loss: 0.1532 - acc: 0.9479 - ETA: 2s - loss: 0.1539 - acc: 0.9479 - ETA: 2s - loss: 0.1543 - acc: 0.9482 - ETA: 2s - loss: 0.1543 - acc: 0.9483 - ETA: 2s - loss: 0.1546 - acc: 0.9478 - ETA: 2s - loss: 0.1542 - acc: 0.9481 - ETA: 2s - loss: 0.1539 - acc: 0.9481 - ETA: 2s - loss: 0.1529 - acc: 0.9481 - ETA: 2s - loss: 0.1556 - acc: 0.9471 - ETA: 2s - loss: 0.1576 - acc: 0.9467 - ETA: 2s - loss: 0.1612 - acc: 0.9462 - ETA: 2s - loss: 0.1602 - acc: 0.9465 - ETA: 1s - loss: 0.1607 - acc: 0.9466 - ETA: 1s - loss: 0.1609 - acc: 0.9464 - ETA: 1s - loss: 0.1598 - acc: 0.9467 - ETA: 1s - loss: 0.1599 - acc: 0.9465 - ETA: 1s - loss: 0.1607 - acc: 0.9459 - ETA: 1s - loss: 0.1615 - acc: 0.9453 - ETA: 1s - loss: 0.1604 - acc: 0.9458 - ETA: 1s - loss: 0.1607 - acc: 0.9457 - ETA: 1s - loss: 0.1627 - acc: 0.9455 - ETA: 1s - loss: 0.1645 - acc: 0.9454 - ETA: 1s - loss: 0.1677 - acc: 0.9450 - ETA: 1s - loss: 0.1678 - acc: 0.9449 - ETA: 1s - loss: 0.1683 - acc: 0.9446 - ETA: 1s - loss: 0.1671 - acc: 0.9451 - ETA: 1s - loss: 0.1688 - acc: 0.9450 - ETA: 1s - loss: 0.1700 - acc: 0.9454 - ETA: 1s - loss: 0.1700 - acc: 0.9455 - ETA: 0s - loss: 0.1684 - acc: 0.9461 - ETA: 0s - loss: 0.1677 - acc: 0.9464 - ETA: 0s - loss: 0.1676 - acc: 0.9466 - ETA: 0s - loss: 0.1668 - acc: 0.9466 - ETA: 0s - loss: 0.1656 - acc: 0.9469 - ETA: 0s - loss: 0.1658 - acc: 0.9467 - ETA: 0s - loss: 0.1644 - acc: 0.9469 - ETA: 0s - loss: 0.1645 - acc: 0.9468 - ETA: 0s - loss: 0.1665 - acc: 0.9464 - ETA: 0s - loss: 0.1653 - acc: 0.9467 - ETA: 0s - loss: 0.1652 - acc: 0.9466 - ETA: 0s - loss: 0.1690 - acc: 0.9459 - ETA: 0s - loss: 0.1702 - acc: 0.9454 - ETA: 0s - loss: 0.1704 - acc: 0.9453 - ETA: 0s - loss: 0.1704 - acc: 0.9452 - ETA: 0s - loss: 0.1704 - acc: 0.9451 - ETA: 0s - loss: 0.1720 - acc: 0.9446Epoch 00006: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.1724 - acc: 0.9445 - val_loss: 0.6832 - val_acc: 0.8515
    Epoch 8/20
    6640/6680 [============================>.] - ETA: 4s - loss: 0.2507 - acc: 0.9000 - ETA: 4s - loss: 0.1212 - acc: 0.9600 - ETA: 4s - loss: 0.0852 - acc: 0.9722 - ETA: 4s - loss: 0.1130 - acc: 0.9692 - ETA: 4s - loss: 0.1334 - acc: 0.9588 - ETA: 4s - loss: 0.1244 - acc: 0.9595 - ETA: 4s - loss: 0.1107 - acc: 0.9620 - ETA: 4s - loss: 0.1191 - acc: 0.9586 - ETA: 4s - loss: 0.1126 - acc: 0.9606 - ETA: 4s - loss: 0.1200 - acc: 0.9595 - ETA: 4s - loss: 0.1119 - acc: 0.9622 - ETA: 4s - loss: 0.1133 - acc: 0.9633 - ETA: 4s - loss: 0.1344 - acc: 0.9561 - ETA: 4s - loss: 0.1314 - acc: 0.9566 - ETA: 4s - loss: 0.1268 - acc: 0.9570 - ETA: 3s - loss: 0.1270 - acc: 0.9574 - ETA: 3s - loss: 0.1236 - acc: 0.9592 - ETA: 3s - loss: 0.1253 - acc: 0.9588 - ETA: 3s - loss: 0.1285 - acc: 0.9592 - ETA: 3s - loss: 0.1279 - acc: 0.9593 - ETA: 3s - loss: 0.1243 - acc: 0.9601 - ETA: 3s - loss: 0.1266 - acc: 0.9596 - ETA: 3s - loss: 0.1263 - acc: 0.9586 - ETA: 3s - loss: 0.1307 - acc: 0.9577 - ETA: 3s - loss: 0.1312 - acc: 0.9574 - ETA: 3s - loss: 0.1354 - acc: 0.9566 - ETA: 3s - loss: 0.1333 - acc: 0.9569 - ETA: 3s - loss: 0.1306 - acc: 0.9580 - ETA: 3s - loss: 0.1308 - acc: 0.9577 - ETA: 3s - loss: 0.1293 - acc: 0.9579 - ETA: 3s - loss: 0.1289 - acc: 0.9577 - ETA: 3s - loss: 0.1311 - acc: 0.9579 - ETA: 3s - loss: 0.1323 - acc: 0.9573 - ETA: 3s - loss: 0.1305 - acc: 0.9570 - ETA: 3s - loss: 0.1282 - acc: 0.9580 - ETA: 3s - loss: 0.1268 - acc: 0.9581 - ETA: 2s - loss: 0.1304 - acc: 0.9571 - ETA: 2s - loss: 0.1304 - acc: 0.9576 - ETA: 2s - loss: 0.1288 - acc: 0.9584 - ETA: 2s - loss: 0.1283 - acc: 0.9586 - ETA: 2s - loss: 0.1288 - acc: 0.9590 - ETA: 2s - loss: 0.1295 - acc: 0.9584 - ETA: 2s - loss: 0.1315 - acc: 0.9576 - ETA: 2s - loss: 0.1303 - acc: 0.9577 - ETA: 2s - loss: 0.1308 - acc: 0.9578 - ETA: 2s - loss: 0.1300 - acc: 0.9577 - ETA: 2s - loss: 0.1306 - acc: 0.9572 - ETA: 2s - loss: 0.1307 - acc: 0.9573 - ETA: 2s - loss: 0.1284 - acc: 0.9582 - ETA: 2s - loss: 0.1323 - acc: 0.9581 - ETA: 2s - loss: 0.1320 - acc: 0.9582 - ETA: 2s - loss: 0.1317 - acc: 0.9577 - ETA: 1s - loss: 0.1364 - acc: 0.9574 - ETA: 1s - loss: 0.1351 - acc: 0.9575 - ETA: 1s - loss: 0.1363 - acc: 0.9568 - ETA: 1s - loss: 0.1381 - acc: 0.9565 - ETA: 1s - loss: 0.1367 - acc: 0.9568 - ETA: 1s - loss: 0.1392 - acc: 0.9562 - ETA: 1s - loss: 0.1409 - acc: 0.9559 - ETA: 1s - loss: 0.1424 - acc: 0.9554 - ETA: 1s - loss: 0.1449 - acc: 0.9551 - ETA: 1s - loss: 0.1443 - acc: 0.9552 - ETA: 1s - loss: 0.1441 - acc: 0.9556 - ETA: 1s - loss: 0.1434 - acc: 0.9555 - ETA: 1s - loss: 0.1442 - acc: 0.9554 - ETA: 1s - loss: 0.1446 - acc: 0.9549 - ETA: 1s - loss: 0.1437 - acc: 0.9550 - ETA: 1s - loss: 0.1440 - acc: 0.9552 - ETA: 1s - loss: 0.1429 - acc: 0.9553 - ETA: 1s - loss: 0.1424 - acc: 0.9554 - ETA: 0s - loss: 0.1435 - acc: 0.9550 - ETA: 0s - loss: 0.1434 - acc: 0.9551 - ETA: 0s - loss: 0.1436 - acc: 0.9548 - ETA: 0s - loss: 0.1434 - acc: 0.9548 - ETA: 0s - loss: 0.1424 - acc: 0.9550 - ETA: 0s - loss: 0.1440 - acc: 0.9548 - ETA: 0s - loss: 0.1449 - acc: 0.9546 - ETA: 0s - loss: 0.1442 - acc: 0.9548 - ETA: 0s - loss: 0.1460 - acc: 0.9544 - ETA: 0s - loss: 0.1461 - acc: 0.9545 - ETA: 0s - loss: 0.1478 - acc: 0.9543 - ETA: 0s - loss: 0.1487 - acc: 0.9541 - ETA: 0s - loss: 0.1479 - acc: 0.9542 - ETA: 0s - loss: 0.1480 - acc: 0.9543 - ETA: 0s - loss: 0.1474 - acc: 0.9546 - ETA: 0s - loss: 0.1472 - acc: 0.9547Epoch 00007: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.1470 - acc: 0.9546 - val_loss: 0.7338 - val_acc: 0.8587
    Epoch 9/20
    6620/6680 [============================>.] - ETA: 4s - loss: 0.0025 - acc: 1.0000 - ETA: 4s - loss: 0.0784 - acc: 0.9800 - ETA: 4s - loss: 0.0907 - acc: 0.9611 - ETA: 4s - loss: 0.1106 - acc: 0.9615 - ETA: 4s - loss: 0.1202 - acc: 0.9559 - ETA: 4s - loss: 0.1219 - acc: 0.9619 - ETA: 4s - loss: 0.1212 - acc: 0.9620 - ETA: 4s - loss: 0.1196 - acc: 0.9638 - ETA: 4s - loss: 0.1175 - acc: 0.9621 - ETA: 4s - loss: 0.1160 - acc: 0.9649 - ETA: 4s - loss: 0.1130 - acc: 0.9659 - ETA: 4s - loss: 0.1084 - acc: 0.9656 - ETA: 4s - loss: 0.1041 - acc: 0.9667 - ETA: 4s - loss: 0.1079 - acc: 0.9663 - ETA: 4s - loss: 0.1098 - acc: 0.9652 - ETA: 4s - loss: 0.1064 - acc: 0.9650 - ETA: 4s - loss: 0.1107 - acc: 0.9648 - ETA: 3s - loss: 0.1119 - acc: 0.9647 - ETA: 3s - loss: 0.1201 - acc: 0.9632 - ETA: 3s - loss: 0.1159 - acc: 0.9645 - ETA: 3s - loss: 0.1194 - acc: 0.9644 - ETA: 3s - loss: 0.1212 - acc: 0.9649 - ETA: 3s - loss: 0.1199 - acc: 0.9653 - ETA: 3s - loss: 0.1238 - acc: 0.9630 - ETA: 3s - loss: 0.1224 - acc: 0.9641 - ETA: 3s - loss: 0.1312 - acc: 0.9625 - ETA: 3s - loss: 0.1284 - acc: 0.9620 - ETA: 3s - loss: 0.1290 - acc: 0.9616 - ETA: 3s - loss: 0.1261 - acc: 0.9621 - ETA: 3s - loss: 0.1224 - acc: 0.9634 - ETA: 3s - loss: 0.1211 - acc: 0.9637 - ETA: 3s - loss: 0.1185 - acc: 0.9645 - ETA: 3s - loss: 0.1216 - acc: 0.9641 - ETA: 3s - loss: 0.1205 - acc: 0.9640 - ETA: 3s - loss: 0.1273 - acc: 0.9629 - ETA: 2s - loss: 0.1260 - acc: 0.9629 - ETA: 2s - loss: 0.1238 - acc: 0.9636 - ETA: 2s - loss: 0.1224 - acc: 0.9639 - ETA: 2s - loss: 0.1223 - acc: 0.9636 - ETA: 2s - loss: 0.1236 - acc: 0.9626 - ETA: 2s - loss: 0.1229 - acc: 0.9623 - ETA: 2s - loss: 0.1208 - acc: 0.9626 - ETA: 2s - loss: 0.1201 - acc: 0.9629 - ETA: 2s - loss: 0.1228 - acc: 0.9623 - ETA: 2s - loss: 0.1238 - acc: 0.9620 - ETA: 2s - loss: 0.1226 - acc: 0.9623 - ETA: 2s - loss: 0.1237 - acc: 0.9620 - ETA: 2s - loss: 0.1231 - acc: 0.9615 - ETA: 2s - loss: 0.1242 - acc: 0.9610 - ETA: 2s - loss: 0.1242 - acc: 0.9608 - ETA: 2s - loss: 0.1262 - acc: 0.9598 - ETA: 2s - loss: 0.1249 - acc: 0.9599 - ETA: 1s - loss: 0.1244 - acc: 0.9600 - ETA: 1s - loss: 0.1229 - acc: 0.9602 - ETA: 1s - loss: 0.1225 - acc: 0.9603 - ETA: 1s - loss: 0.1215 - acc: 0.9604 - ETA: 1s - loss: 0.1223 - acc: 0.9602 - ETA: 1s - loss: 0.1221 - acc: 0.9604 - ETA: 1s - loss: 0.1223 - acc: 0.9607 - ETA: 1s - loss: 0.1224 - acc: 0.9612 - ETA: 1s - loss: 0.1225 - acc: 0.9608 - ETA: 1s - loss: 0.1229 - acc: 0.9604 - ETA: 1s - loss: 0.1220 - acc: 0.9606 - ETA: 1s - loss: 0.1217 - acc: 0.9608 - ETA: 1s - loss: 0.1214 - acc: 0.9611 - ETA: 1s - loss: 0.1216 - acc: 0.9611 - ETA: 1s - loss: 0.1211 - acc: 0.9613 - ETA: 1s - loss: 0.1206 - acc: 0.9614 - ETA: 1s - loss: 0.1205 - acc: 0.9616 - ETA: 0s - loss: 0.1204 - acc: 0.9613 - ETA: 0s - loss: 0.1195 - acc: 0.9615 - ETA: 0s - loss: 0.1239 - acc: 0.9608 - ETA: 0s - loss: 0.1249 - acc: 0.9604 - ETA: 0s - loss: 0.1257 - acc: 0.9599 - ETA: 0s - loss: 0.1248 - acc: 0.9600 - ETA: 0s - loss: 0.1245 - acc: 0.9602 - ETA: 0s - loss: 0.1253 - acc: 0.9599 - ETA: 0s - loss: 0.1266 - acc: 0.9592 - ETA: 0s - loss: 0.1260 - acc: 0.9594 - ETA: 0s - loss: 0.1262 - acc: 0.9598 - ETA: 0s - loss: 0.1259 - acc: 0.9598 - ETA: 0s - loss: 0.1247 - acc: 0.9602 - ETA: 0s - loss: 0.1250 - acc: 0.9602 - ETA: 0s - loss: 0.1246 - acc: 0.9604 - ETA: 0s - loss: 0.1246 - acc: 0.9603Epoch 00008: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.1247 - acc: 0.9605 - val_loss: 0.7781 - val_acc: 0.8383
    Epoch 10/20
    6640/6680 [============================>.] - ETA: 4s - loss: 0.1805 - acc: 0.9500 - ETA: 4s - loss: 0.0577 - acc: 0.9900 - ETA: 4s - loss: 0.0640 - acc: 0.9833 - ETA: 4s - loss: 0.0993 - acc: 0.9731 - ETA: 4s - loss: 0.0933 - acc: 0.9765 - ETA: 4s - loss: 0.1002 - acc: 0.9786 - ETA: 4s - loss: 0.0921 - acc: 0.9780 - ETA: 4s - loss: 0.0943 - acc: 0.9759 - ETA: 4s - loss: 0.0981 - acc: 0.9758 - ETA: 4s - loss: 0.1040 - acc: 0.9743 - ETA: 4s - loss: 0.0979 - acc: 0.9756 - ETA: 4s - loss: 0.0956 - acc: 0.9767 - ETA: 4s - loss: 0.0903 - acc: 0.9776 - ETA: 4s - loss: 0.0862 - acc: 0.9783 - ETA: 4s - loss: 0.0877 - acc: 0.9781 - ETA: 3s - loss: 0.0906 - acc: 0.9762 - ETA: 3s - loss: 0.0899 - acc: 0.9769 - ETA: 3s - loss: 0.0882 - acc: 0.9761 - ETA: 3s - loss: 0.0895 - acc: 0.9760 - ETA: 3s - loss: 0.0964 - acc: 0.9740 - ETA: 3s - loss: 0.0957 - acc: 0.9741 - ETA: 3s - loss: 0.0953 - acc: 0.9735 - ETA: 3s - loss: 0.0993 - acc: 0.9733 - ETA: 3s - loss: 0.0992 - acc: 0.9725 - ETA: 3s - loss: 0.0995 - acc: 0.9726 - ETA: 3s - loss: 0.1057 - acc: 0.9712 - ETA: 3s - loss: 0.1050 - acc: 0.9709 - ETA: 3s - loss: 0.1033 - acc: 0.9715 - ETA: 3s - loss: 0.1041 - acc: 0.9712 - ETA: 3s - loss: 0.1018 - acc: 0.9722 - ETA: 3s - loss: 0.1015 - acc: 0.9714 - ETA: 3s - loss: 0.1042 - acc: 0.9699 - ETA: 3s - loss: 0.1015 - acc: 0.9709 - ETA: 3s - loss: 0.1021 - acc: 0.9702 - ETA: 3s - loss: 0.1009 - acc: 0.9704 - ETA: 2s - loss: 0.1013 - acc: 0.9691 - ETA: 2s - loss: 0.0994 - acc: 0.9697 - ETA: 2s - loss: 0.0988 - acc: 0.9697 - ETA: 2s - loss: 0.1040 - acc: 0.9693 - ETA: 2s - loss: 0.1023 - acc: 0.9697 - ETA: 2s - loss: 0.1021 - acc: 0.9697 - ETA: 2s - loss: 0.1008 - acc: 0.9698 - ETA: 2s - loss: 0.0998 - acc: 0.9702 - ETA: 2s - loss: 0.0993 - acc: 0.9705 - ETA: 2s - loss: 0.1001 - acc: 0.9703 - ETA: 2s - loss: 0.0987 - acc: 0.9704 - ETA: 2s - loss: 0.0975 - acc: 0.9708 - ETA: 2s - loss: 0.0991 - acc: 0.9703 - ETA: 2s - loss: 0.0989 - acc: 0.9699 - ETA: 2s - loss: 0.0993 - acc: 0.9695 - ETA: 2s - loss: 0.0977 - acc: 0.9701 - ETA: 2s - loss: 0.0969 - acc: 0.9702 - ETA: 2s - loss: 0.0969 - acc: 0.9700 - ETA: 1s - loss: 0.0975 - acc: 0.9694 - ETA: 1s - loss: 0.0977 - acc: 0.9688 - ETA: 1s - loss: 0.0973 - acc: 0.9689 - ETA: 1s - loss: 0.0970 - acc: 0.9688 - ETA: 1s - loss: 0.0973 - acc: 0.9687 - ETA: 1s - loss: 0.0975 - acc: 0.9684 - ETA: 1s - loss: 0.0988 - acc: 0.9680 - ETA: 1s - loss: 0.0999 - acc: 0.9679 - ETA: 1s - loss: 0.1003 - acc: 0.9681 - ETA: 1s - loss: 0.1034 - acc: 0.9678 - ETA: 1s - loss: 0.1030 - acc: 0.9678 - ETA: 1s - loss: 0.1065 - acc: 0.9671 - ETA: 1s - loss: 0.1065 - acc: 0.9670 - ETA: 1s - loss: 0.1076 - acc: 0.9667 - ETA: 1s - loss: 0.1081 - acc: 0.9670 - ETA: 1s - loss: 0.1080 - acc: 0.9674 - ETA: 1s - loss: 0.1069 - acc: 0.9678 - ETA: 0s - loss: 0.1056 - acc: 0.9683 - ETA: 0s - loss: 0.1048 - acc: 0.9684 - ETA: 0s - loss: 0.1050 - acc: 0.9681 - ETA: 0s - loss: 0.1053 - acc: 0.9681 - ETA: 0s - loss: 0.1063 - acc: 0.9675 - ETA: 0s - loss: 0.1084 - acc: 0.9674 - ETA: 0s - loss: 0.1076 - acc: 0.9675 - ETA: 0s - loss: 0.1064 - acc: 0.9679 - ETA: 0s - loss: 0.1072 - acc: 0.9674 - ETA: 0s - loss: 0.1071 - acc: 0.9672 - ETA: 0s - loss: 0.1082 - acc: 0.9669 - ETA: 0s - loss: 0.1079 - acc: 0.9670 - ETA: 0s - loss: 0.1078 - acc: 0.9670 - ETA: 0s - loss: 0.1075 - acc: 0.9671 - ETA: 0s - loss: 0.1077 - acc: 0.9671 - ETA: 0s - loss: 0.1070 - acc: 0.9673Epoch 00009: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.1073 - acc: 0.9671 - val_loss: 0.7530 - val_acc: 0.8611
    Epoch 11/20
    6600/6680 [============================>.] - ETA: 4s - loss: 0.0244 - acc: 1.0000 - ETA: 4s - loss: 0.0807 - acc: 0.9900 - ETA: 4s - loss: 0.0716 - acc: 0.9889 - ETA: 4s - loss: 0.0729 - acc: 0.9846 - ETA: 4s - loss: 0.0676 - acc: 0.9853 - ETA: 4s - loss: 0.0787 - acc: 0.9786 - ETA: 4s - loss: 0.0708 - acc: 0.9800 - ETA: 4s - loss: 0.0835 - acc: 0.9776 - ETA: 4s - loss: 0.0856 - acc: 0.9758 - ETA: 4s - loss: 0.0792 - acc: 0.9770 - ETA: 4s - loss: 0.0739 - acc: 0.9780 - ETA: 4s - loss: 0.0734 - acc: 0.9767 - ETA: 4s - loss: 0.0771 - acc: 0.9745 - ETA: 3s - loss: 0.0828 - acc: 0.9717 - ETA: 3s - loss: 0.0786 - acc: 0.9737 - ETA: 3s - loss: 0.0872 - acc: 0.9721 - ETA: 3s - loss: 0.0850 - acc: 0.9731 - ETA: 3s - loss: 0.0824 - acc: 0.9739 - ETA: 3s - loss: 0.0790 - acc: 0.9747 - ETA: 3s - loss: 0.0788 - acc: 0.9747 - ETA: 3s - loss: 0.0758 - acc: 0.9759 - ETA: 3s - loss: 0.0729 - acc: 0.9771 - ETA: 3s - loss: 0.0737 - acc: 0.9770 - ETA: 3s - loss: 0.0718 - acc: 0.9780 - ETA: 3s - loss: 0.0712 - acc: 0.9784 - ETA: 3s - loss: 0.0697 - acc: 0.9787 - ETA: 3s - loss: 0.0742 - acc: 0.9786 - ETA: 3s - loss: 0.0748 - acc: 0.9775 - ETA: 3s - loss: 0.0759 - acc: 0.9774 - ETA: 3s - loss: 0.0740 - acc: 0.9782 - ETA: 3s - loss: 0.0787 - acc: 0.9756 - ETA: 2s - loss: 0.0767 - acc: 0.9764 - ETA: 2s - loss: 0.0764 - acc: 0.9760 - ETA: 2s - loss: 0.0756 - acc: 0.9759 - ETA: 2s - loss: 0.0748 - acc: 0.9763 - ETA: 2s - loss: 0.0734 - acc: 0.9770 - ETA: 2s - loss: 0.0741 - acc: 0.9759 - ETA: 2s - loss: 0.0759 - acc: 0.9755 - ETA: 2s - loss: 0.0772 - acc: 0.9752 - ETA: 2s - loss: 0.0772 - acc: 0.9752 - ETA: 2s - loss: 0.0811 - acc: 0.9739 - ETA: 2s - loss: 0.0830 - acc: 0.9727 - ETA: 2s - loss: 0.0819 - acc: 0.9728 - ETA: 2s - loss: 0.0826 - acc: 0.9725 - ETA: 2s - loss: 0.0851 - acc: 0.9720 - ETA: 2s - loss: 0.0835 - acc: 0.9727 - ETA: 2s - loss: 0.0827 - acc: 0.9727 - ETA: 2s - loss: 0.0832 - acc: 0.9730 - ETA: 2s - loss: 0.0842 - acc: 0.9728 - ETA: 1s - loss: 0.0850 - acc: 0.9721 - ETA: 1s - loss: 0.0835 - acc: 0.9726 - ETA: 1s - loss: 0.0830 - acc: 0.9728 - ETA: 1s - loss: 0.0826 - acc: 0.9731 - ETA: 1s - loss: 0.0819 - acc: 0.9731 - ETA: 1s - loss: 0.0832 - acc: 0.9729 - ETA: 1s - loss: 0.0830 - acc: 0.9730 - ETA: 1s - loss: 0.0846 - acc: 0.9725 - ETA: 1s - loss: 0.0850 - acc: 0.9726 - ETA: 1s - loss: 0.0839 - acc: 0.9731 - ETA: 1s - loss: 0.0834 - acc: 0.9729 - ETA: 1s - loss: 0.0843 - acc: 0.9729 - ETA: 1s - loss: 0.0839 - acc: 0.9730 - ETA: 1s - loss: 0.0843 - acc: 0.9728 - ETA: 1s - loss: 0.0845 - acc: 0.9726 - ETA: 1s - loss: 0.0835 - acc: 0.9730 - ETA: 1s - loss: 0.0834 - acc: 0.9731 - ETA: 1s - loss: 0.0840 - acc: 0.9729 - ETA: 0s - loss: 0.0851 - acc: 0.9728 - ETA: 0s - loss: 0.0884 - acc: 0.9722 - ETA: 0s - loss: 0.0887 - acc: 0.9721 - ETA: 0s - loss: 0.0897 - acc: 0.9718 - ETA: 0s - loss: 0.0888 - acc: 0.9722 - ETA: 0s - loss: 0.0898 - acc: 0.9722 - ETA: 0s - loss: 0.0901 - acc: 0.9721 - ETA: 0s - loss: 0.0913 - acc: 0.9718 - ETA: 0s - loss: 0.0905 - acc: 0.9721 - ETA: 0s - loss: 0.0899 - acc: 0.9723 - ETA: 0s - loss: 0.0891 - acc: 0.9725 - ETA: 0s - loss: 0.0894 - acc: 0.9727 - ETA: 0s - loss: 0.0886 - acc: 0.9730 - ETA: 0s - loss: 0.0879 - acc: 0.9732 - ETA: 0s - loss: 0.0904 - acc: 0.9728 - ETA: 0s - loss: 0.0899 - acc: 0.9729 - ETA: 0s - loss: 0.0906 - acc: 0.9726Epoch 00010: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0922 - acc: 0.9720 - val_loss: 0.8565 - val_acc: 0.8503
    Epoch 12/20
    6640/6680 [============================>.] - ETA: 4s - loss: 0.0250 - acc: 1.0000 - ETA: 4s - loss: 0.0195 - acc: 0.9900 - ETA: 4s - loss: 0.0373 - acc: 0.9833 - ETA: 4s - loss: 0.0477 - acc: 0.9769 - ETA: 4s - loss: 0.0509 - acc: 0.9735 - ETA: 4s - loss: 0.0871 - acc: 0.9690 - ETA: 4s - loss: 0.0776 - acc: 0.9720 - ETA: 4s - loss: 0.0732 - acc: 0.9724 - ETA: 4s - loss: 0.0673 - acc: 0.9742 - ETA: 4s - loss: 0.0647 - acc: 0.9757 - ETA: 4s - loss: 0.0652 - acc: 0.9756 - ETA: 4s - loss: 0.0611 - acc: 0.9778 - ETA: 4s - loss: 0.0675 - acc: 0.9745 - ETA: 4s - loss: 0.0678 - acc: 0.9745 - ETA: 4s - loss: 0.0640 - acc: 0.9763 - ETA: 4s - loss: 0.0646 - acc: 0.9754 - ETA: 4s - loss: 0.0678 - acc: 0.9746 - ETA: 4s - loss: 0.0650 - acc: 0.9761 - ETA: 4s - loss: 0.0701 - acc: 0.9753 - ETA: 3s - loss: 0.0672 - acc: 0.9766 - ETA: 3s - loss: 0.0684 - acc: 0.9747 - ETA: 3s - loss: 0.0706 - acc: 0.9753 - ETA: 3s - loss: 0.0694 - acc: 0.9756 - ETA: 3s - loss: 0.0736 - acc: 0.9750 - ETA: 3s - loss: 0.0773 - acc: 0.9734 - ETA: 3s - loss: 0.0754 - acc: 0.9735 - ETA: 3s - loss: 0.0734 - acc: 0.9740 - ETA: 3s - loss: 0.0749 - acc: 0.9736 - ETA: 3s - loss: 0.0748 - acc: 0.9737 - ETA: 3s - loss: 0.0727 - acc: 0.9746 - ETA: 3s - loss: 0.0714 - acc: 0.9746 - ETA: 3s - loss: 0.0750 - acc: 0.9734 - ETA: 3s - loss: 0.0747 - acc: 0.9730 - ETA: 3s - loss: 0.0753 - acc: 0.9735 - ETA: 3s - loss: 0.0752 - acc: 0.9735 - ETA: 2s - loss: 0.0740 - acc: 0.9739 - ETA: 2s - loss: 0.0726 - acc: 0.9747 - ETA: 2s - loss: 0.0721 - acc: 0.9750 - ETA: 2s - loss: 0.0713 - acc: 0.9750 - ETA: 2s - loss: 0.0700 - acc: 0.9753 - ETA: 2s - loss: 0.0700 - acc: 0.9750 - ETA: 2s - loss: 0.0692 - acc: 0.9753 - ETA: 2s - loss: 0.0698 - acc: 0.9750 - ETA: 2s - loss: 0.0699 - acc: 0.9750 - ETA: 2s - loss: 0.0707 - acc: 0.9744 - ETA: 2s - loss: 0.0722 - acc: 0.9742 - ETA: 2s - loss: 0.0711 - acc: 0.9745 - ETA: 2s - loss: 0.0721 - acc: 0.9739 - ETA: 2s - loss: 0.0711 - acc: 0.9742 - ETA: 2s - loss: 0.0699 - acc: 0.9747 - ETA: 2s - loss: 0.0696 - acc: 0.9747 - ETA: 1s - loss: 0.0714 - acc: 0.9745 - ETA: 1s - loss: 0.0708 - acc: 0.9748 - ETA: 1s - loss: 0.0705 - acc: 0.9750 - ETA: 1s - loss: 0.0712 - acc: 0.9752 - ETA: 1s - loss: 0.0729 - acc: 0.9745 - ETA: 1s - loss: 0.0739 - acc: 0.9746 - ETA: 1s - loss: 0.0757 - acc: 0.9739 - ETA: 1s - loss: 0.0754 - acc: 0.9744 - ETA: 1s - loss: 0.0774 - acc: 0.9744 - ETA: 1s - loss: 0.0771 - acc: 0.9742 - ETA: 1s - loss: 0.0786 - acc: 0.9742 - ETA: 1s - loss: 0.0781 - acc: 0.9740 - ETA: 1s - loss: 0.0788 - acc: 0.9734 - ETA: 1s - loss: 0.0780 - acc: 0.9738 - ETA: 1s - loss: 0.0796 - acc: 0.9738 - ETA: 1s - loss: 0.0787 - acc: 0.9742 - ETA: 0s - loss: 0.0788 - acc: 0.9744 - ETA: 0s - loss: 0.0779 - acc: 0.9748 - ETA: 0s - loss: 0.0777 - acc: 0.9746 - ETA: 0s - loss: 0.0772 - acc: 0.9746 - ETA: 0s - loss: 0.0769 - acc: 0.9745 - ETA: 0s - loss: 0.0764 - acc: 0.9747 - ETA: 0s - loss: 0.0759 - acc: 0.9748 - ETA: 0s - loss: 0.0780 - acc: 0.9743 - ETA: 0s - loss: 0.0779 - acc: 0.9745 - ETA: 0s - loss: 0.0772 - acc: 0.9747 - ETA: 0s - loss: 0.0771 - acc: 0.9748 - ETA: 0s - loss: 0.0769 - acc: 0.9747 - ETA: 0s - loss: 0.0762 - acc: 0.9750 - ETA: 0s - loss: 0.0756 - acc: 0.9752 - ETA: 0s - loss: 0.0758 - acc: 0.9750 - ETA: 0s - loss: 0.0765 - acc: 0.9752 - ETA: 0s - loss: 0.0766 - acc: 0.9752Epoch 00011: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0768 - acc: 0.9749 - val_loss: 0.8031 - val_acc: 0.8575
    Epoch 13/20
    6600/6680 [============================>.] - ETA: 5s - loss: 0.0849 - acc: 0.9000 - ETA: 4s - loss: 0.0623 - acc: 0.9700 - ETA: 5s - loss: 0.0515 - acc: 0.9750 - ETA: 5s - loss: 0.0525 - acc: 0.9727 - ETA: 5s - loss: 0.0509 - acc: 0.9750 - ETA: 5s - loss: 0.0428 - acc: 0.9794 - ETA: 5s - loss: 0.0417 - acc: 0.9810 - ETA: 5s - loss: 0.0491 - acc: 0.9800 - ETA: 5s - loss: 0.0560 - acc: 0.9786 - ETA: 5s - loss: 0.0508 - acc: 0.9806 - ETA: 4s - loss: 0.0475 - acc: 0.9824 - ETA: 4s - loss: 0.0543 - acc: 0.9811 - ETA: 4s - loss: 0.0507 - acc: 0.9825 - ETA: 4s - loss: 0.0526 - acc: 0.9830 - ETA: 4s - loss: 0.0563 - acc: 0.9830 - ETA: 4s - loss: 0.0629 - acc: 0.9804 - ETA: 4s - loss: 0.0595 - acc: 0.9818 - ETA: 4s - loss: 0.0587 - acc: 0.9819 - ETA: 4s - loss: 0.0619 - acc: 0.9803 - ETA: 4s - loss: 0.0602 - acc: 0.9808 - ETA: 4s - loss: 0.0647 - acc: 0.9801 - ETA: 4s - loss: 0.0643 - acc: 0.9803 - ETA: 4s - loss: 0.0654 - acc: 0.9791 - ETA: 4s - loss: 0.0634 - acc: 0.9799 - ETA: 4s - loss: 0.0609 - acc: 0.9809 - ETA: 4s - loss: 0.0593 - acc: 0.9812 - ETA: 4s - loss: 0.0590 - acc: 0.9815 - ETA: 3s - loss: 0.0604 - acc: 0.9812 - ETA: 3s - loss: 0.0630 - acc: 0.9809 - ETA: 3s - loss: 0.0608 - acc: 0.9817 - ETA: 3s - loss: 0.0599 - acc: 0.9819 - ETA: 3s - loss: 0.0606 - acc: 0.9817 - ETA: 3s - loss: 0.0590 - acc: 0.9823 - ETA: 3s - loss: 0.0587 - acc: 0.9821 - ETA: 3s - loss: 0.0586 - acc: 0.9818 - ETA: 3s - loss: 0.0575 - acc: 0.9824 - ETA: 3s - loss: 0.0614 - acc: 0.9818 - ETA: 3s - loss: 0.0608 - acc: 0.9823 - ETA: 3s - loss: 0.0613 - acc: 0.9824 - ETA: 3s - loss: 0.0601 - acc: 0.9827 - ETA: 3s - loss: 0.0595 - acc: 0.9825 - ETA: 3s - loss: 0.0587 - acc: 0.9827 - ETA: 2s - loss: 0.0582 - acc: 0.9828 - ETA: 2s - loss: 0.0574 - acc: 0.9829 - ETA: 2s - loss: 0.0564 - acc: 0.9830 - ETA: 2s - loss: 0.0609 - acc: 0.9822 - ETA: 2s - loss: 0.0605 - acc: 0.9823 - ETA: 2s - loss: 0.0599 - acc: 0.9824 - ETA: 2s - loss: 0.0596 - acc: 0.9825 - ETA: 2s - loss: 0.0602 - acc: 0.9820 - ETA: 2s - loss: 0.0624 - acc: 0.9816 - ETA: 2s - loss: 0.0615 - acc: 0.9817 - ETA: 2s - loss: 0.0618 - acc: 0.9818 - ETA: 2s - loss: 0.0616 - acc: 0.9817 - ETA: 2s - loss: 0.0632 - acc: 0.9813 - ETA: 2s - loss: 0.0630 - acc: 0.9814 - ETA: 2s - loss: 0.0649 - acc: 0.9808 - ETA: 1s - loss: 0.0657 - acc: 0.9807 - ETA: 1s - loss: 0.0650 - acc: 0.9808 - ETA: 1s - loss: 0.0666 - acc: 0.9805 - ETA: 1s - loss: 0.0670 - acc: 0.9804 - ETA: 1s - loss: 0.0683 - acc: 0.9803 - ETA: 1s - loss: 0.0696 - acc: 0.9802 - ETA: 1s - loss: 0.0693 - acc: 0.9803 - ETA: 1s - loss: 0.0707 - acc: 0.9797 - ETA: 1s - loss: 0.0721 - acc: 0.9793 - ETA: 1s - loss: 0.0717 - acc: 0.9793 - ETA: 1s - loss: 0.0714 - acc: 0.9794 - ETA: 1s - loss: 0.0710 - acc: 0.9793 - ETA: 1s - loss: 0.0710 - acc: 0.9794 - ETA: 1s - loss: 0.0711 - acc: 0.9793 - ETA: 1s - loss: 0.0711 - acc: 0.9793 - ETA: 1s - loss: 0.0702 - acc: 0.9796 - ETA: 1s - loss: 0.0698 - acc: 0.9797 - ETA: 0s - loss: 0.0713 - acc: 0.9795 - ETA: 0s - loss: 0.0722 - acc: 0.9792 - ETA: 0s - loss: 0.0727 - acc: 0.9792 - ETA: 0s - loss: 0.0719 - acc: 0.9794 - ETA: 0s - loss: 0.0713 - acc: 0.9796 - ETA: 0s - loss: 0.0718 - acc: 0.9793 - ETA: 0s - loss: 0.0722 - acc: 0.9790 - ETA: 0s - loss: 0.0724 - acc: 0.9788 - ETA: 0s - loss: 0.0728 - acc: 0.9784 - ETA: 0s - loss: 0.0721 - acc: 0.9787 - ETA: 0s - loss: 0.0720 - acc: 0.9785 - ETA: 0s - loss: 0.0722 - acc: 0.9783 - ETA: 0s - loss: 0.0716 - acc: 0.9784 - ETA: 0s - loss: 0.0723 - acc: 0.9782 - ETA: 0s - loss: 0.0719 - acc: 0.9782Epoch 00012: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0731 - acc: 0.9781 - val_loss: 0.7755 - val_acc: 0.8647
    Epoch 14/20
    6660/6680 [============================>.] - ETA: 4s - loss: 0.1340 - acc: 0.9000 - ETA: 4s - loss: 0.0348 - acc: 0.9800 - ETA: 4s - loss: 0.0644 - acc: 0.9778 - ETA: 4s - loss: 0.0545 - acc: 0.9808 - ETA: 4s - loss: 0.0771 - acc: 0.9794 - ETA: 4s - loss: 0.0838 - acc: 0.9762 - ETA: 4s - loss: 0.0733 - acc: 0.9800 - ETA: 4s - loss: 0.0674 - acc: 0.9810 - ETA: 4s - loss: 0.0683 - acc: 0.9803 - ETA: 4s - loss: 0.0667 - acc: 0.9811 - ETA: 4s - loss: 0.0729 - acc: 0.9793 - ETA: 4s - loss: 0.0706 - acc: 0.9789 - ETA: 4s - loss: 0.0653 - acc: 0.9806 - ETA: 3s - loss: 0.0618 - acc: 0.9821 - ETA: 3s - loss: 0.0635 - acc: 0.9816 - ETA: 3s - loss: 0.0602 - acc: 0.9828 - ETA: 3s - loss: 0.0603 - acc: 0.9831 - ETA: 3s - loss: 0.0661 - acc: 0.9826 - ETA: 3s - loss: 0.0636 - acc: 0.9836 - ETA: 3s - loss: 0.0660 - acc: 0.9831 - ETA: 3s - loss: 0.0629 - acc: 0.9840 - ETA: 3s - loss: 0.0629 - acc: 0.9835 - ETA: 3s - loss: 0.0606 - acc: 0.9843 - ETA: 3s - loss: 0.0603 - acc: 0.9844 - ETA: 3s - loss: 0.0611 - acc: 0.9845 - ETA: 3s - loss: 0.0604 - acc: 0.9847 - ETA: 3s - loss: 0.0598 - acc: 0.9843 - ETA: 3s - loss: 0.0660 - acc: 0.9830 - ETA: 3s - loss: 0.0651 - acc: 0.9832 - ETA: 3s - loss: 0.0661 - acc: 0.9829 - ETA: 3s - loss: 0.0653 - acc: 0.9826 - ETA: 2s - loss: 0.0644 - acc: 0.9824 - ETA: 2s - loss: 0.0644 - acc: 0.9822 - ETA: 2s - loss: 0.0638 - acc: 0.9823 - ETA: 2s - loss: 0.0682 - acc: 0.9821 - ETA: 2s - loss: 0.0669 - acc: 0.9826 - ETA: 2s - loss: 0.0656 - acc: 0.9828 - ETA: 2s - loss: 0.0663 - acc: 0.9829 - ETA: 2s - loss: 0.0650 - acc: 0.9833 - ETA: 2s - loss: 0.0634 - acc: 0.9838 - ETA: 2s - loss: 0.0625 - acc: 0.9842 - ETA: 2s - loss: 0.0622 - acc: 0.9842 - ETA: 2s - loss: 0.0610 - acc: 0.9846 - ETA: 2s - loss: 0.0598 - acc: 0.9850 - ETA: 2s - loss: 0.0600 - acc: 0.9847 - ETA: 2s - loss: 0.0595 - acc: 0.9848 - ETA: 2s - loss: 0.0595 - acc: 0.9846 - ETA: 2s - loss: 0.0589 - acc: 0.9849 - ETA: 2s - loss: 0.0581 - acc: 0.9850 - ETA: 1s - loss: 0.0572 - acc: 0.9853 - ETA: 1s - loss: 0.0564 - acc: 0.9856 - ETA: 1s - loss: 0.0572 - acc: 0.9851 - ETA: 1s - loss: 0.0590 - acc: 0.9842 - ETA: 1s - loss: 0.0609 - acc: 0.9840 - ETA: 1s - loss: 0.0613 - acc: 0.9839 - ETA: 1s - loss: 0.0625 - acc: 0.9833 - ETA: 1s - loss: 0.0619 - acc: 0.9836 - ETA: 1s - loss: 0.0629 - acc: 0.9834 - ETA: 1s - loss: 0.0633 - acc: 0.9835 - ETA: 1s - loss: 0.0625 - acc: 0.9838 - ETA: 1s - loss: 0.0617 - acc: 0.9838 - ETA: 1s - loss: 0.0610 - acc: 0.9841 - ETA: 1s - loss: 0.0608 - acc: 0.9841 - ETA: 1s - loss: 0.0618 - acc: 0.9838 - ETA: 1s - loss: 0.0614 - acc: 0.9837 - ETA: 1s - loss: 0.0612 - acc: 0.9837 - ETA: 0s - loss: 0.0614 - acc: 0.9836 - ETA: 0s - loss: 0.0615 - acc: 0.9835 - ETA: 0s - loss: 0.0619 - acc: 0.9833 - ETA: 0s - loss: 0.0622 - acc: 0.9832 - ETA: 0s - loss: 0.0619 - acc: 0.9831 - ETA: 0s - loss: 0.0618 - acc: 0.9832 - ETA: 0s - loss: 0.0626 - acc: 0.9829 - ETA: 0s - loss: 0.0625 - acc: 0.9826 - ETA: 0s - loss: 0.0632 - acc: 0.9823 - ETA: 0s - loss: 0.0625 - acc: 0.9826 - ETA: 0s - loss: 0.0620 - acc: 0.9826 - ETA: 0s - loss: 0.0618 - acc: 0.9825 - ETA: 0s - loss: 0.0614 - acc: 0.9826 - ETA: 0s - loss: 0.0607 - acc: 0.9828 - ETA: 0s - loss: 0.0612 - acc: 0.9827 - ETA: 0s - loss: 0.0607 - acc: 0.9829 - ETA: 0s - loss: 0.0605 - acc: 0.9830 - ETA: 0s - loss: 0.0617 - acc: 0.9826Epoch 00013: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0615 - acc: 0.9826 - val_loss: 0.8708 - val_acc: 0.8479
    Epoch 15/20
    6640/6680 [============================>.] - ETA: 5s - loss: 0.0390 - acc: 1.0000 - ETA: 5s - loss: 0.0207 - acc: 1.0000 - ETA: 5s - loss: 0.0136 - acc: 1.0000 - ETA: 5s - loss: 0.0119 - acc: 1.0000 - ETA: 5s - loss: 0.0165 - acc: 0.9937 - ETA: 4s - loss: 0.0152 - acc: 0.9950 - ETA: 4s - loss: 0.0153 - acc: 0.9958 - ETA: 4s - loss: 0.0202 - acc: 0.9946 - ETA: 4s - loss: 0.0389 - acc: 0.9906 - ETA: 4s - loss: 0.0359 - acc: 0.9917 - ETA: 4s - loss: 0.0343 - acc: 0.9925 - ETA: 4s - loss: 0.0321 - acc: 0.9932 - ETA: 4s - loss: 0.0311 - acc: 0.9937 - ETA: 4s - loss: 0.0310 - acc: 0.9923 - ETA: 4s - loss: 0.0362 - acc: 0.9911 - ETA: 4s - loss: 0.0357 - acc: 0.9908 - ETA: 4s - loss: 0.0343 - acc: 0.9914 - ETA: 4s - loss: 0.0343 - acc: 0.9910 - ETA: 4s - loss: 0.0334 - acc: 0.9914 - ETA: 4s - loss: 0.0396 - acc: 0.9892 - ETA: 4s - loss: 0.0472 - acc: 0.9878 - ETA: 3s - loss: 0.0463 - acc: 0.9872 - ETA: 3s - loss: 0.0445 - acc: 0.9878 - ETA: 3s - loss: 0.0460 - acc: 0.9872 - ETA: 3s - loss: 0.0483 - acc: 0.9867 - ETA: 3s - loss: 0.0478 - acc: 0.9867 - ETA: 3s - loss: 0.0464 - acc: 0.9873 - ETA: 3s - loss: 0.0469 - acc: 0.9873 - ETA: 3s - loss: 0.0457 - acc: 0.9877 - ETA: 3s - loss: 0.0449 - acc: 0.9877 - ETA: 3s - loss: 0.0435 - acc: 0.9881 - ETA: 3s - loss: 0.0452 - acc: 0.9881 - ETA: 3s - loss: 0.0441 - acc: 0.9885 - ETA: 3s - loss: 0.0434 - acc: 0.9888 - ETA: 3s - loss: 0.0438 - acc: 0.9883 - ETA: 3s - loss: 0.0433 - acc: 0.9883 - ETA: 3s - loss: 0.0425 - acc: 0.9887 - ETA: 2s - loss: 0.0471 - acc: 0.9883 - ETA: 2s - loss: 0.0462 - acc: 0.9886 - ETA: 2s - loss: 0.0451 - acc: 0.9889 - ETA: 2s - loss: 0.0443 - acc: 0.9892 - ETA: 2s - loss: 0.0480 - acc: 0.9879 - ETA: 2s - loss: 0.0470 - acc: 0.9882 - ETA: 2s - loss: 0.0462 - acc: 0.9885 - ETA: 2s - loss: 0.0471 - acc: 0.9884 - ETA: 2s - loss: 0.0464 - acc: 0.9884 - ETA: 2s - loss: 0.0456 - acc: 0.9887 - ETA: 2s - loss: 0.0478 - acc: 0.9876 - ETA: 2s - loss: 0.0476 - acc: 0.9876 - ETA: 2s - loss: 0.0484 - acc: 0.9870 - ETA: 2s - loss: 0.0511 - acc: 0.9862 - ETA: 2s - loss: 0.0538 - acc: 0.9855 - ETA: 2s - loss: 0.0537 - acc: 0.9855 - ETA: 1s - loss: 0.0545 - acc: 0.9856 - ETA: 1s - loss: 0.0553 - acc: 0.9856 - ETA: 1s - loss: 0.0550 - acc: 0.9856 - ETA: 1s - loss: 0.0556 - acc: 0.9855 - ETA: 1s - loss: 0.0558 - acc: 0.9853 - ETA: 1s - loss: 0.0552 - acc: 0.9855 - ETA: 1s - loss: 0.0552 - acc: 0.9855 - ETA: 1s - loss: 0.0545 - acc: 0.9857 - ETA: 1s - loss: 0.0539 - acc: 0.9860 - ETA: 1s - loss: 0.0539 - acc: 0.9860 - ETA: 1s - loss: 0.0540 - acc: 0.9856 - ETA: 1s - loss: 0.0544 - acc: 0.9857 - ETA: 1s - loss: 0.0555 - acc: 0.9853 - ETA: 1s - loss: 0.0548 - acc: 0.9855 - ETA: 1s - loss: 0.0546 - acc: 0.9855 - ETA: 1s - loss: 0.0543 - acc: 0.9853 - ETA: 1s - loss: 0.0538 - acc: 0.9856 - ETA: 0s - loss: 0.0536 - acc: 0.9854 - ETA: 0s - loss: 0.0540 - acc: 0.9853 - ETA: 0s - loss: 0.0548 - acc: 0.9848 - ETA: 0s - loss: 0.0547 - acc: 0.9848 - ETA: 0s - loss: 0.0550 - acc: 0.9847 - ETA: 0s - loss: 0.0544 - acc: 0.9849 - ETA: 0s - loss: 0.0555 - acc: 0.9847 - ETA: 0s - loss: 0.0551 - acc: 0.9847 - ETA: 0s - loss: 0.0554 - acc: 0.9843 - ETA: 0s - loss: 0.0557 - acc: 0.9841 - ETA: 0s - loss: 0.0559 - acc: 0.9840 - ETA: 0s - loss: 0.0552 - acc: 0.9842 - ETA: 0s - loss: 0.0548 - acc: 0.9844 - ETA: 0s - loss: 0.0542 - acc: 0.9846 - ETA: 0s - loss: 0.0543 - acc: 0.9845 - ETA: 0s - loss: 0.0538 - acc: 0.9846Epoch 00014: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0537 - acc: 0.9846 - val_loss: 0.8716 - val_acc: 0.8515
    Epoch 16/20
    6660/6680 [============================>.] - ETA: 4s - loss: 0.0075 - acc: 1.0000 - ETA: 4s - loss: 0.0248 - acc: 0.9900 - ETA: 4s - loss: 0.0533 - acc: 0.9833 - ETA: 4s - loss: 0.0417 - acc: 0.9885 - ETA: 4s - loss: 0.0421 - acc: 0.9882 - ETA: 4s - loss: 0.0366 - acc: 0.9905 - ETA: 4s - loss: 0.0399 - acc: 0.9860 - ETA: 4s - loss: 0.0398 - acc: 0.9845 - ETA: 4s - loss: 0.0434 - acc: 0.9833 - ETA: 4s - loss: 0.0411 - acc: 0.9851 - ETA: 4s - loss: 0.0400 - acc: 0.9841 - ETA: 4s - loss: 0.0382 - acc: 0.9856 - ETA: 4s - loss: 0.0361 - acc: 0.9867 - ETA: 4s - loss: 0.0381 - acc: 0.9858 - ETA: 3s - loss: 0.0365 - acc: 0.9860 - ETA: 3s - loss: 0.0368 - acc: 0.9861 - ETA: 3s - loss: 0.0350 - acc: 0.9869 - ETA: 3s - loss: 0.0345 - acc: 0.9870 - ETA: 3s - loss: 0.0331 - acc: 0.9877 - ETA: 3s - loss: 0.0369 - acc: 0.9877 - ETA: 3s - loss: 0.0356 - acc: 0.9883 - ETA: 3s - loss: 0.0366 - acc: 0.9871 - ETA: 3s - loss: 0.0352 - acc: 0.9876 - ETA: 3s - loss: 0.0387 - acc: 0.9866 - ETA: 3s - loss: 0.0372 - acc: 0.9871 - ETA: 3s - loss: 0.0364 - acc: 0.9876 - ETA: 3s - loss: 0.0358 - acc: 0.9881 - ETA: 3s - loss: 0.0348 - acc: 0.9885 - ETA: 3s - loss: 0.0346 - acc: 0.9885 - ETA: 3s - loss: 0.0337 - acc: 0.9889 - ETA: 3s - loss: 0.0336 - acc: 0.9888 - ETA: 3s - loss: 0.0341 - acc: 0.9888 - ETA: 2s - loss: 0.0340 - acc: 0.9888 - ETA: 2s - loss: 0.0355 - acc: 0.9887 - ETA: 2s - loss: 0.0359 - acc: 0.9883 - ETA: 2s - loss: 0.0364 - acc: 0.9883 - ETA: 2s - loss: 0.0368 - acc: 0.9883 - ETA: 2s - loss: 0.0362 - acc: 0.9886 - ETA: 2s - loss: 0.0359 - acc: 0.9886 - ETA: 2s - loss: 0.0352 - acc: 0.9889 - ETA: 2s - loss: 0.0376 - acc: 0.9885 - ETA: 2s - loss: 0.0386 - acc: 0.9882 - ETA: 2s - loss: 0.0380 - acc: 0.9885 - ETA: 2s - loss: 0.0401 - acc: 0.9879 - ETA: 2s - loss: 0.0401 - acc: 0.9879 - ETA: 2s - loss: 0.0402 - acc: 0.9876 - ETA: 2s - loss: 0.0425 - acc: 0.9873 - ETA: 2s - loss: 0.0418 - acc: 0.9876 - ETA: 2s - loss: 0.0419 - acc: 0.9876 - ETA: 1s - loss: 0.0442 - acc: 0.9873 - ETA: 1s - loss: 0.0435 - acc: 0.9876 - ETA: 1s - loss: 0.0456 - acc: 0.9866 - ETA: 1s - loss: 0.0469 - acc: 0.9861 - ETA: 1s - loss: 0.0464 - acc: 0.9862 - ETA: 1s - loss: 0.0460 - acc: 0.9864 - ETA: 1s - loss: 0.0454 - acc: 0.9867 - ETA: 1s - loss: 0.0447 - acc: 0.9869 - ETA: 1s - loss: 0.0447 - acc: 0.9869 - ETA: 1s - loss: 0.0445 - acc: 0.9869 - ETA: 1s - loss: 0.0459 - acc: 0.9865 - ETA: 1s - loss: 0.0455 - acc: 0.9867 - ETA: 1s - loss: 0.0457 - acc: 0.9865 - ETA: 1s - loss: 0.0452 - acc: 0.9867 - ETA: 1s - loss: 0.0456 - acc: 0.9866 - ETA: 1s - loss: 0.0463 - acc: 0.9864 - ETA: 1s - loss: 0.0465 - acc: 0.9862 - ETA: 0s - loss: 0.0467 - acc: 0.9860 - ETA: 0s - loss: 0.0463 - acc: 0.9862 - ETA: 0s - loss: 0.0463 - acc: 0.9863 - ETA: 0s - loss: 0.0470 - acc: 0.9857 - ETA: 0s - loss: 0.0466 - acc: 0.9858 - ETA: 0s - loss: 0.0462 - acc: 0.9860 - ETA: 0s - loss: 0.0497 - acc: 0.9856 - ETA: 0s - loss: 0.0491 - acc: 0.9858 - ETA: 0s - loss: 0.0500 - acc: 0.9854 - ETA: 0s - loss: 0.0494 - acc: 0.9855 - ETA: 0s - loss: 0.0491 - acc: 0.9856 - ETA: 0s - loss: 0.0489 - acc: 0.9856 - ETA: 0s - loss: 0.0490 - acc: 0.9855 - ETA: 0s - loss: 0.0489 - acc: 0.9855 - ETA: 0s - loss: 0.0486 - acc: 0.9855 - ETA: 0s - loss: 0.0482 - acc: 0.9857 - ETA: 0s - loss: 0.0477 - acc: 0.9859 - ETA: 0s - loss: 0.0477 - acc: 0.9857Epoch 00015: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0476 - acc: 0.9858 - val_loss: 0.8739 - val_acc: 0.8551
    Epoch 17/20
    6620/6680 [============================>.] - ETA: 4s - loss: 0.0603 - acc: 0.9500 - ETA: 4s - loss: 0.0211 - acc: 0.9900 - ETA: 4s - loss: 0.0160 - acc: 0.9944 - ETA: 4s - loss: 0.0123 - acc: 0.9962 - ETA: 4s - loss: 0.0138 - acc: 0.9941 - ETA: 4s - loss: 0.0132 - acc: 0.9952 - ETA: 4s - loss: 0.0142 - acc: 0.9940 - ETA: 4s - loss: 0.0135 - acc: 0.9948 - ETA: 4s - loss: 0.0156 - acc: 0.9939 - ETA: 4s - loss: 0.0213 - acc: 0.9919 - ETA: 4s - loss: 0.0239 - acc: 0.9902 - ETA: 4s - loss: 0.0229 - acc: 0.9911 - ETA: 4s - loss: 0.0250 - acc: 0.9908 - ETA: 4s - loss: 0.0278 - acc: 0.9896 - ETA: 4s - loss: 0.0264 - acc: 0.9904 - ETA: 3s - loss: 0.0284 - acc: 0.9893 - ETA: 3s - loss: 0.0337 - acc: 0.9869 - ETA: 3s - loss: 0.0342 - acc: 0.9870 - ETA: 3s - loss: 0.0335 - acc: 0.9877 - ETA: 3s - loss: 0.0325 - acc: 0.9883 - ETA: 3s - loss: 0.0330 - acc: 0.9877 - ETA: 3s - loss: 0.0339 - acc: 0.9876 - ETA: 3s - loss: 0.0345 - acc: 0.9876 - ETA: 3s - loss: 0.0356 - acc: 0.9876 - ETA: 3s - loss: 0.0349 - acc: 0.9875 - ETA: 3s - loss: 0.0358 - acc: 0.9874 - ETA: 3s - loss: 0.0375 - acc: 0.9869 - ETA: 3s - loss: 0.0386 - acc: 0.9860 - ETA: 3s - loss: 0.0383 - acc: 0.9856 - ETA: 3s - loss: 0.0386 - acc: 0.9857 - ETA: 3s - loss: 0.0393 - acc: 0.9853 - ETA: 3s - loss: 0.0384 - acc: 0.9858 - ETA: 3s - loss: 0.0375 - acc: 0.9862 - ETA: 3s - loss: 0.0365 - acc: 0.9866 - ETA: 3s - loss: 0.0364 - acc: 0.9867 - ETA: 2s - loss: 0.0358 - acc: 0.9867 - ETA: 2s - loss: 0.0350 - acc: 0.9871 - ETA: 2s - loss: 0.0353 - acc: 0.9867 - ETA: 2s - loss: 0.0345 - acc: 0.9871 - ETA: 2s - loss: 0.0347 - acc: 0.9868 - ETA: 2s - loss: 0.0374 - acc: 0.9855 - ETA: 2s - loss: 0.0371 - acc: 0.9855 - ETA: 2s - loss: 0.0370 - acc: 0.9855 - ETA: 2s - loss: 0.0362 - acc: 0.9858 - ETA: 2s - loss: 0.0356 - acc: 0.9861 - ETA: 2s - loss: 0.0351 - acc: 0.9864 - ETA: 2s - loss: 0.0345 - acc: 0.9867 - ETA: 2s - loss: 0.0342 - acc: 0.9868 - ETA: 2s - loss: 0.0350 - acc: 0.9868 - ETA: 2s - loss: 0.0352 - acc: 0.9868 - ETA: 2s - loss: 0.0346 - acc: 0.9871 - ETA: 2s - loss: 0.0343 - acc: 0.9871 - ETA: 1s - loss: 0.0337 - acc: 0.9873 - ETA: 1s - loss: 0.0344 - acc: 0.9873 - ETA: 1s - loss: 0.0342 - acc: 0.9872 - ETA: 1s - loss: 0.0351 - acc: 0.9870 - ETA: 1s - loss: 0.0361 - acc: 0.9870 - ETA: 1s - loss: 0.0357 - acc: 0.9870 - ETA: 1s - loss: 0.0355 - acc: 0.9872 - ETA: 1s - loss: 0.0353 - acc: 0.9873 - ETA: 1s - loss: 0.0355 - acc: 0.9873 - ETA: 1s - loss: 0.0351 - acc: 0.9874 - ETA: 1s - loss: 0.0347 - acc: 0.9877 - ETA: 1s - loss: 0.0344 - acc: 0.9877 - ETA: 1s - loss: 0.0341 - acc: 0.9878 - ETA: 1s - loss: 0.0342 - acc: 0.9878 - ETA: 1s - loss: 0.0349 - acc: 0.9876 - ETA: 1s - loss: 0.0346 - acc: 0.9877 - ETA: 1s - loss: 0.0354 - acc: 0.9875 - ETA: 1s - loss: 0.0361 - acc: 0.9871 - ETA: 1s - loss: 0.0415 - acc: 0.9861 - ETA: 0s - loss: 0.0418 - acc: 0.9859 - ETA: 0s - loss: 0.0419 - acc: 0.9858 - ETA: 0s - loss: 0.0416 - acc: 0.9860 - ETA: 0s - loss: 0.0411 - acc: 0.9862 - ETA: 0s - loss: 0.0407 - acc: 0.9863 - ETA: 0s - loss: 0.0404 - acc: 0.9865 - ETA: 0s - loss: 0.0406 - acc: 0.9863 - ETA: 0s - loss: 0.0405 - acc: 0.9861 - ETA: 0s - loss: 0.0401 - acc: 0.9863 - ETA: 0s - loss: 0.0399 - acc: 0.9863 - ETA: 0s - loss: 0.0398 - acc: 0.9864 - ETA: 0s - loss: 0.0394 - acc: 0.9865 - ETA: 0s - loss: 0.0411 - acc: 0.9862 - ETA: 0s - loss: 0.0408 - acc: 0.9864 - ETA: 0s - loss: 0.0413 - acc: 0.9859 - ETA: 0s - loss: 0.0412 - acc: 0.9858 - ETA: 0s - loss: 0.0415 - acc: 0.9853Epoch 00016: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0412 - acc: 0.9855 - val_loss: 0.8735 - val_acc: 0.8551
    Epoch 18/20
    6640/6680 [============================>.] - ETA: 4s - loss: 0.0030 - acc: 1.0000 - ETA: 4s - loss: 0.0190 - acc: 0.9900 - ETA: 4s - loss: 0.0296 - acc: 0.9889 - ETA: 4s - loss: 0.0228 - acc: 0.9923 - ETA: 4s - loss: 0.0273 - acc: 0.9912 - ETA: 4s - loss: 0.0359 - acc: 0.9905 - ETA: 4s - loss: 0.0307 - acc: 0.9920 - ETA: 4s - loss: 0.0284 - acc: 0.9914 - ETA: 4s - loss: 0.0265 - acc: 0.9909 - ETA: 4s - loss: 0.0241 - acc: 0.9919 - ETA: 4s - loss: 0.0228 - acc: 0.9927 - ETA: 4s - loss: 0.0224 - acc: 0.9932 - ETA: 4s - loss: 0.0233 - acc: 0.9927 - ETA: 4s - loss: 0.0260 - acc: 0.9913 - ETA: 4s - loss: 0.0250 - acc: 0.9920 - ETA: 4s - loss: 0.0248 - acc: 0.9917 - ETA: 4s - loss: 0.0248 - acc: 0.9922 - ETA: 4s - loss: 0.0242 - acc: 0.9926 - ETA: 4s - loss: 0.0246 - acc: 0.9917 - ETA: 4s - loss: 0.0244 - acc: 0.9921 - ETA: 3s - loss: 0.0233 - acc: 0.9925 - ETA: 3s - loss: 0.0229 - acc: 0.9923 - ETA: 3s - loss: 0.0224 - acc: 0.9926 - ETA: 3s - loss: 0.0224 - acc: 0.9929 - ETA: 3s - loss: 0.0226 - acc: 0.9927 - ETA: 3s - loss: 0.0229 - acc: 0.9925 - ETA: 3s - loss: 0.0242 - acc: 0.9918 - ETA: 3s - loss: 0.0254 - acc: 0.9917 - ETA: 3s - loss: 0.0251 - acc: 0.9920 - ETA: 3s - loss: 0.0257 - acc: 0.9914 - ETA: 3s - loss: 0.0264 - acc: 0.9912 - ETA: 3s - loss: 0.0265 - acc: 0.9911 - ETA: 3s - loss: 0.0258 - acc: 0.9914 - ETA: 3s - loss: 0.0252 - acc: 0.9917 - ETA: 3s - loss: 0.0271 - acc: 0.9915 - ETA: 2s - loss: 0.0279 - acc: 0.9911 - ETA: 2s - loss: 0.0302 - acc: 0.9906 - ETA: 2s - loss: 0.0315 - acc: 0.9902 - ETA: 2s - loss: 0.0315 - acc: 0.9898 - ETA: 2s - loss: 0.0330 - acc: 0.9894 - ETA: 2s - loss: 0.0324 - acc: 0.9897 - ETA: 2s - loss: 0.0317 - acc: 0.9899 - ETA: 2s - loss: 0.0320 - acc: 0.9899 - ETA: 2s - loss: 0.0321 - acc: 0.9895 - ETA: 2s - loss: 0.0328 - acc: 0.9889 - ETA: 2s - loss: 0.0334 - acc: 0.9889 - ETA: 2s - loss: 0.0337 - acc: 0.9889 - ETA: 2s - loss: 0.0351 - acc: 0.9886 - ETA: 2s - loss: 0.0346 - acc: 0.9888 - ETA: 2s - loss: 0.0342 - acc: 0.9888 - ETA: 2s - loss: 0.0339 - acc: 0.9887 - ETA: 1s - loss: 0.0333 - acc: 0.9890 - ETA: 1s - loss: 0.0348 - acc: 0.9887 - ETA: 1s - loss: 0.0352 - acc: 0.9884 - ETA: 1s - loss: 0.0350 - acc: 0.9887 - ETA: 1s - loss: 0.0349 - acc: 0.9889 - ETA: 1s - loss: 0.0346 - acc: 0.9888 - ETA: 1s - loss: 0.0344 - acc: 0.9888 - ETA: 1s - loss: 0.0338 - acc: 0.9890 - ETA: 1s - loss: 0.0334 - acc: 0.9892 - ETA: 1s - loss: 0.0336 - acc: 0.9890 - ETA: 1s - loss: 0.0337 - acc: 0.9889 - ETA: 1s - loss: 0.0338 - acc: 0.9889 - ETA: 1s - loss: 0.0335 - acc: 0.9889 - ETA: 1s - loss: 0.0331 - acc: 0.9891 - ETA: 1s - loss: 0.0348 - acc: 0.9885 - ETA: 1s - loss: 0.0345 - acc: 0.9886 - ETA: 0s - loss: 0.0351 - acc: 0.9884 - ETA: 0s - loss: 0.0348 - acc: 0.9886 - ETA: 0s - loss: 0.0346 - acc: 0.9886 - ETA: 0s - loss: 0.0345 - acc: 0.9886 - ETA: 0s - loss: 0.0351 - acc: 0.9884 - ETA: 0s - loss: 0.0349 - acc: 0.9884 - ETA: 0s - loss: 0.0345 - acc: 0.9885 - ETA: 0s - loss: 0.0348 - acc: 0.9885 - ETA: 0s - loss: 0.0346 - acc: 0.9887 - ETA: 0s - loss: 0.0346 - acc: 0.9887 - ETA: 0s - loss: 0.0344 - acc: 0.9888 - ETA: 0s - loss: 0.0341 - acc: 0.9889 - ETA: 0s - loss: 0.0339 - acc: 0.9891 - ETA: 0s - loss: 0.0341 - acc: 0.9889 - ETA: 0s - loss: 0.0344 - acc: 0.9889 - ETA: 0s - loss: 0.0348 - acc: 0.9889 - ETA: 0s - loss: 0.0347 - acc: 0.9890Epoch 00017: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0346 - acc: 0.9891 - val_loss: 0.9533 - val_acc: 0.8515
    Epoch 19/20
    6620/6680 [============================>.] - ETA: 4s - loss: 8.9530e-04 - acc: 1.0000 - ETA: 4s - loss: 0.0325 - acc: 0.9900     - ETA: 4s - loss: 0.0197 - acc: 0.9944 - ETA: 4s - loss: 0.0172 - acc: 0.9923 - ETA: 4s - loss: 0.0168 - acc: 0.9941 - ETA: 4s - loss: 0.0140 - acc: 0.9952 - ETA: 4s - loss: 0.0120 - acc: 0.9960 - ETA: 4s - loss: 0.0164 - acc: 0.9948 - ETA: 4s - loss: 0.0177 - acc: 0.9939 - ETA: 4s - loss: 0.0162 - acc: 0.9946 - ETA: 4s - loss: 0.0200 - acc: 0.9927 - ETA: 4s - loss: 0.0186 - acc: 0.9933 - ETA: 4s - loss: 0.0179 - acc: 0.9939 - ETA: 4s - loss: 0.0176 - acc: 0.9934 - ETA: 3s - loss: 0.0165 - acc: 0.9939 - ETA: 3s - loss: 0.0184 - acc: 0.9926 - ETA: 3s - loss: 0.0177 - acc: 0.9931 - ETA: 3s - loss: 0.0179 - acc: 0.9928 - ETA: 3s - loss: 0.0172 - acc: 0.9932 - ETA: 3s - loss: 0.0191 - acc: 0.9922 - ETA: 3s - loss: 0.0188 - acc: 0.9926 - ETA: 3s - loss: 0.0192 - acc: 0.9924 - ETA: 3s - loss: 0.0186 - acc: 0.9927 - ETA: 3s - loss: 0.0202 - acc: 0.9919 - ETA: 3s - loss: 0.0195 - acc: 0.9923 - ETA: 3s - loss: 0.0198 - acc: 0.9921 - ETA: 3s - loss: 0.0202 - acc: 0.9919 - ETA: 3s - loss: 0.0218 - acc: 0.9917 - ETA: 3s - loss: 0.0222 - acc: 0.9916 - ETA: 3s - loss: 0.0224 - acc: 0.9915 - ETA: 3s - loss: 0.0222 - acc: 0.9913 - ETA: 3s - loss: 0.0217 - acc: 0.9916 - ETA: 2s - loss: 0.0225 - acc: 0.9915 - ETA: 2s - loss: 0.0224 - acc: 0.9914 - ETA: 2s - loss: 0.0250 - acc: 0.9909 - ETA: 2s - loss: 0.0257 - acc: 0.9908 - ETA: 2s - loss: 0.0258 - acc: 0.9907 - ETA: 2s - loss: 0.0254 - acc: 0.9909 - ETA: 2s - loss: 0.0251 - acc: 0.9908 - ETA: 2s - loss: 0.0254 - acc: 0.9908 - ETA: 2s - loss: 0.0252 - acc: 0.9910 - ETA: 2s - loss: 0.0246 - acc: 0.9912 - ETA: 2s - loss: 0.0246 - acc: 0.9911 - ETA: 2s - loss: 0.0243 - acc: 0.9913 - ETA: 2s - loss: 0.0238 - acc: 0.9915 - ETA: 2s - loss: 0.0266 - acc: 0.9912 - ETA: 2s - loss: 0.0280 - acc: 0.9905 - ETA: 2s - loss: 0.0276 - acc: 0.9907 - ETA: 2s - loss: 0.0271 - acc: 0.9909 - ETA: 1s - loss: 0.0272 - acc: 0.9906 - ETA: 1s - loss: 0.0290 - acc: 0.9903 - ETA: 1s - loss: 0.0287 - acc: 0.9902 - ETA: 1s - loss: 0.0286 - acc: 0.9902 - ETA: 1s - loss: 0.0281 - acc: 0.9904 - ETA: 1s - loss: 0.0278 - acc: 0.9906 - ETA: 1s - loss: 0.0283 - acc: 0.9905 - ETA: 1s - loss: 0.0285 - acc: 0.9904 - ETA: 1s - loss: 0.0284 - acc: 0.9904 - ETA: 1s - loss: 0.0279 - acc: 0.9906 - ETA: 1s - loss: 0.0289 - acc: 0.9903 - ETA: 1s - loss: 0.0296 - acc: 0.9898 - ETA: 1s - loss: 0.0292 - acc: 0.9900 - ETA: 1s - loss: 0.0292 - acc: 0.9900 - ETA: 1s - loss: 0.0317 - acc: 0.9897 - ETA: 1s - loss: 0.0314 - acc: 0.9898 - ETA: 1s - loss: 0.0312 - acc: 0.9900 - ETA: 1s - loss: 0.0318 - acc: 0.9896 - ETA: 0s - loss: 0.0333 - acc: 0.9894 - ETA: 0s - loss: 0.0334 - acc: 0.9890 - ETA: 0s - loss: 0.0346 - acc: 0.9886 - ETA: 0s - loss: 0.0346 - acc: 0.9884 - ETA: 0s - loss: 0.0352 - acc: 0.9882 - ETA: 0s - loss: 0.0350 - acc: 0.9882 - ETA: 0s - loss: 0.0349 - acc: 0.9882 - ETA: 0s - loss: 0.0347 - acc: 0.9883 - ETA: 0s - loss: 0.0344 - acc: 0.9885 - ETA: 0s - loss: 0.0346 - acc: 0.9885 - ETA: 0s - loss: 0.0349 - acc: 0.9883 - ETA: 0s - loss: 0.0345 - acc: 0.9885 - ETA: 0s - loss: 0.0346 - acc: 0.9884 - ETA: 0s - loss: 0.0344 - acc: 0.9884 - ETA: 0s - loss: 0.0341 - acc: 0.9885 - ETA: 0s - loss: 0.0340 - acc: 0.9885 - ETA: 0s - loss: 0.0339 - acc: 0.9885Epoch 00018: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0339 - acc: 0.9885 - val_loss: 0.9198 - val_acc: 0.8491
    Epoch 20/20
    6660/6680 [============================>.] - ETA: 4s - loss: 0.0630 - acc: 0.9500 - ETA: 4s - loss: 0.0221 - acc: 0.9900 - ETA: 4s - loss: 0.0650 - acc: 0.9889 - ETA: 4s - loss: 0.0548 - acc: 0.9885 - ETA: 4s - loss: 0.0480 - acc: 0.9882 - ETA: 4s - loss: 0.0466 - acc: 0.9881 - ETA: 4s - loss: 0.0411 - acc: 0.9900 - ETA: 4s - loss: 0.0363 - acc: 0.9914 - ETA: 4s - loss: 0.0345 - acc: 0.9906 - ETA: 4s - loss: 0.0310 - acc: 0.9917 - ETA: 4s - loss: 0.0285 - acc: 0.9925 - ETA: 4s - loss: 0.0277 - acc: 0.9920 - ETA: 4s - loss: 0.0269 - acc: 0.9917 - ETA: 4s - loss: 0.0251 - acc: 0.9923 - ETA: 4s - loss: 0.0249 - acc: 0.9911 - ETA: 3s - loss: 0.0245 - acc: 0.9908 - ETA: 3s - loss: 0.0235 - acc: 0.9914 - ETA: 3s - loss: 0.0232 - acc: 0.9912 - ETA: 3s - loss: 0.0274 - acc: 0.9910 - ETA: 3s - loss: 0.0295 - acc: 0.9908 - ETA: 3s - loss: 0.0286 - acc: 0.9912 - ETA: 3s - loss: 0.0281 - acc: 0.9917 - ETA: 3s - loss: 0.0271 - acc: 0.9920 - ETA: 3s - loss: 0.0263 - acc: 0.9924 - ETA: 3s - loss: 0.0257 - acc: 0.9927 - ETA: 3s - loss: 0.0253 - acc: 0.9925 - ETA: 3s - loss: 0.0267 - acc: 0.9918 - ETA: 3s - loss: 0.0263 - acc: 0.9921 - ETA: 3s - loss: 0.0255 - acc: 0.9924 - ETA: 3s - loss: 0.0259 - acc: 0.9922 - ETA: 3s - loss: 0.0254 - acc: 0.9925 - ETA: 3s - loss: 0.0247 - acc: 0.9927 - ETA: 2s - loss: 0.0247 - acc: 0.9926 - ETA: 2s - loss: 0.0243 - acc: 0.9928 - ETA: 2s - loss: 0.0241 - acc: 0.9930 - ETA: 2s - loss: 0.0237 - acc: 0.9932 - ETA: 2s - loss: 0.0239 - acc: 0.9931 - ETA: 2s - loss: 0.0238 - acc: 0.9932 - ETA: 2s - loss: 0.0243 - acc: 0.9931 - ETA: 2s - loss: 0.0262 - acc: 0.9926 - ETA: 2s - loss: 0.0259 - acc: 0.9925 - ETA: 2s - loss: 0.0264 - acc: 0.9924 - ETA: 2s - loss: 0.0258 - acc: 0.9926 - ETA: 2s - loss: 0.0254 - acc: 0.9927 - ETA: 2s - loss: 0.0269 - acc: 0.9926 - ETA: 2s - loss: 0.0301 - acc: 0.9919 - ETA: 2s - loss: 0.0300 - acc: 0.9918 - ETA: 2s - loss: 0.0294 - acc: 0.9920 - ETA: 2s - loss: 0.0289 - acc: 0.9922 - ETA: 1s - loss: 0.0288 - acc: 0.9921 - ETA: 1s - loss: 0.0296 - acc: 0.9920 - ETA: 1s - loss: 0.0317 - acc: 0.9919 - ETA: 1s - loss: 0.0316 - acc: 0.9918 - ETA: 1s - loss: 0.0314 - acc: 0.9920 - ETA: 1s - loss: 0.0316 - acc: 0.9917 - ETA: 1s - loss: 0.0312 - acc: 0.9918 - ETA: 1s - loss: 0.0310 - acc: 0.9917 - ETA: 1s - loss: 0.0310 - acc: 0.9917 - ETA: 1s - loss: 0.0314 - acc: 0.9914 - ETA: 1s - loss: 0.0312 - acc: 0.9913 - ETA: 1s - loss: 0.0312 - acc: 0.9912 - ETA: 1s - loss: 0.0312 - acc: 0.9914 - ETA: 1s - loss: 0.0308 - acc: 0.9915 - ETA: 1s - loss: 0.0305 - acc: 0.9914 - ETA: 1s - loss: 0.0318 - acc: 0.9912 - ETA: 1s - loss: 0.0315 - acc: 0.9913 - ETA: 1s - loss: 0.0312 - acc: 0.9914 - ETA: 0s - loss: 0.0312 - acc: 0.9914 - ETA: 0s - loss: 0.0317 - acc: 0.9913 - ETA: 0s - loss: 0.0324 - acc: 0.9911 - ETA: 0s - loss: 0.0320 - acc: 0.9912 - ETA: 0s - loss: 0.0324 - acc: 0.9910 - ETA: 0s - loss: 0.0322 - acc: 0.9909 - ETA: 0s - loss: 0.0318 - acc: 0.9911 - ETA: 0s - loss: 0.0328 - acc: 0.9908 - ETA: 0s - loss: 0.0326 - acc: 0.9908 - ETA: 0s - loss: 0.0328 - acc: 0.9907 - ETA: 0s - loss: 0.0328 - acc: 0.9907 - ETA: 0s - loss: 0.0325 - acc: 0.9908 - ETA: 0s - loss: 0.0321 - acc: 0.9909 - ETA: 0s - loss: 0.0320 - acc: 0.9909 - ETA: 0s - loss: 0.0317 - acc: 0.9910 - ETA: 0s - loss: 0.0314 - acc: 0.9911 - ETA: 0s - loss: 0.0314 - acc: 0.9910 - ETA: 0s - loss: 0.0311 - acc: 0.9911Epoch 00019: val_loss did not improve
    6680/6680 [==============================] - 5s - loss: 0.0311 - acc: 0.9912 - val_loss: 0.9812 - val_acc: 0.8551
    ---I am done saving model valid_InceptionV3 ----
    
    In [41]:
    ### TODO: Train the model.
    checkpointer_Xception = ModelCheckpoint(filepath='weights.best.Xception.hdf5', 
                                   verbose=1, save_best_only=True)
    
    Xception_model.fit(train_Xception, train_targets, 
              validation_data=(valid_Xception, valid_targets),
              epochs=20, batch_size=20, callbacks=[checkpointer_Xception], verbose=1)
    
    print('---I am done saving model valid_Xception  ----')
    
    Train on 6680 samples, validate on 835 samples
    Epoch 1/20
    6620/6680 [============================>.] - ETA: 274s - loss: 4.8607 - acc: 0.0500 - ETA: 97s - loss: 4.9696 - acc: 0.0333  - ETA: 61s - loss: 4.9393 - acc: 0.0700 - ETA: 53s - loss: 4.8943 - acc: 0.0750 - ETA: 42s - loss: 4.7621 - acc: 0.1000 - ETA: 35s - loss: 4.7242 - acc: 0.1100 - ETA: 31s - loss: 4.6249 - acc: 0.1250 - ETA: 27s - loss: 4.5341 - acc: 0.1536 - ETA: 25s - loss: 4.4944 - acc: 0.1719 - ETA: 23s - loss: 4.3514 - acc: 0.1972 - ETA: 21s - loss: 4.2249 - acc: 0.2200 - ETA: 20s - loss: 4.1353 - acc: 0.2318 - ETA: 19s - loss: 4.0066 - acc: 0.2479 - ETA: 18s - loss: 3.9159 - acc: 0.2558 - ETA: 17s - loss: 3.8220 - acc: 0.2696 - ETA: 16s - loss: 3.6664 - acc: 0.2919 - ETA: 16s - loss: 3.5555 - acc: 0.3121 - ETA: 15s - loss: 3.4820 - acc: 0.3229 - ETA: 15s - loss: 3.3946 - acc: 0.3405 - ETA: 14s - loss: 3.3260 - acc: 0.3513 - ETA: 14s - loss: 3.2398 - acc: 0.3671 - ETA: 14s - loss: 3.1632 - acc: 0.3826 - ETA: 13s - loss: 3.0538 - acc: 0.4022 - ETA: 13s - loss: 3.0241 - acc: 0.4064 - ETA: 13s - loss: 2.9682 - acc: 0.4173 - ETA: 13s - loss: 2.9049 - acc: 0.4225 - ETA: 12s - loss: 2.8516 - acc: 0.4321 - ETA: 12s - loss: 2.7875 - acc: 0.4482 - ETA: 12s - loss: 2.7076 - acc: 0.4603 - ETA: 12s - loss: 2.6659 - acc: 0.4642 - ETA: 11s - loss: 2.6361 - acc: 0.4694 - ETA: 11s - loss: 2.5742 - acc: 0.4800 - ETA: 11s - loss: 2.5362 - acc: 0.4866 - ETA: 11s - loss: 2.5182 - acc: 0.4890 - ETA: 11s - loss: 2.4822 - acc: 0.4943 - ETA: 10s - loss: 2.4440 - acc: 0.5014 - ETA: 10s - loss: 2.3999 - acc: 0.5081 - ETA: 10s - loss: 2.3604 - acc: 0.5138 - ETA: 10s - loss: 2.3257 - acc: 0.5186 - ETA: 10s - loss: 2.2907 - acc: 0.5244 - ETA: 9s - loss: 2.2390 - acc: 0.5337  - ETA: 9s - loss: 2.2109 - acc: 0.5382 - ETA: 9s - loss: 2.1810 - acc: 0.5437 - ETA: 9s - loss: 2.1493 - acc: 0.5494 - ETA: 9s - loss: 2.1225 - acc: 0.5544 - ETA: 9s - loss: 2.1082 - acc: 0.5571 - ETA: 9s - loss: 2.0899 - acc: 0.5596 - ETA: 9s - loss: 2.0735 - acc: 0.5594 - ETA: 8s - loss: 2.0363 - acc: 0.5662 - ETA: 8s - loss: 2.0101 - acc: 0.5708 - ETA: 8s - loss: 1.9849 - acc: 0.5748 - ETA: 8s - loss: 1.9637 - acc: 0.5776 - ETA: 8s - loss: 1.9436 - acc: 0.5799 - ETA: 8s - loss: 1.9221 - acc: 0.5830 - ETA: 8s - loss: 1.8895 - acc: 0.5893 - ETA: 8s - loss: 1.8698 - acc: 0.5925 - ETA: 7s - loss: 1.8429 - acc: 0.5979 - ETA: 7s - loss: 1.8223 - acc: 0.6017 - ETA: 7s - loss: 1.8038 - acc: 0.6037 - ETA: 7s - loss: 1.7882 - acc: 0.6061 - ETA: 7s - loss: 1.7606 - acc: 0.6115 - ETA: 7s - loss: 1.7347 - acc: 0.6163 - ETA: 7s - loss: 1.7214 - acc: 0.6187 - ETA: 7s - loss: 1.7083 - acc: 0.6207 - ETA: 7s - loss: 1.7021 - acc: 0.6216 - ETA: 7s - loss: 1.6871 - acc: 0.6250 - ETA: 7s - loss: 1.6775 - acc: 0.6264 - ETA: 6s - loss: 1.6635 - acc: 0.6293 - ETA: 6s - loss: 1.6498 - acc: 0.6317 - ETA: 6s - loss: 1.6369 - acc: 0.6337 - ETA: 6s - loss: 1.6203 - acc: 0.6370 - ETA: 6s - loss: 1.6062 - acc: 0.6399 - ETA: 6s - loss: 1.5941 - acc: 0.6420 - ETA: 6s - loss: 1.5761 - acc: 0.6451 - ETA: 6s - loss: 1.5662 - acc: 0.6468 - ETA: 6s - loss: 1.5607 - acc: 0.6474 - ETA: 6s - loss: 1.5501 - acc: 0.6487 - ETA: 6s - loss: 1.5428 - acc: 0.6491 - ETA: 6s - loss: 1.5351 - acc: 0.6500 - ETA: 5s - loss: 1.5263 - acc: 0.6509 - ETA: 5s - loss: 1.5161 - acc: 0.6527 - ETA: 5s - loss: 1.5071 - acc: 0.6545 - ETA: 5s - loss: 1.5023 - acc: 0.6550 - ETA: 5s - loss: 1.4914 - acc: 0.6570 - ETA: 5s - loss: 1.4806 - acc: 0.6592 - ETA: 5s - loss: 1.4681 - acc: 0.6624 - ETA: 5s - loss: 1.4583 - acc: 0.6642 - ETA: 5s - loss: 1.4453 - acc: 0.6668 - ETA: 5s - loss: 1.4366 - acc: 0.6688 - ETA: 4s - loss: 1.4255 - acc: 0.6712 - ETA: 4s - loss: 1.4133 - acc: 0.6733 - ETA: 4s - loss: 1.4100 - acc: 0.6734 - ETA: 4s - loss: 1.4062 - acc: 0.6741 - ETA: 4s - loss: 1.4008 - acc: 0.6753 - ETA: 4s - loss: 1.3920 - acc: 0.6768 - ETA: 4s - loss: 1.3851 - acc: 0.6776 - ETA: 4s - loss: 1.3791 - acc: 0.6788 - ETA: 4s - loss: 1.3690 - acc: 0.6813 - ETA: 4s - loss: 1.3630 - acc: 0.6824 - ETA: 4s - loss: 1.3540 - acc: 0.6841 - ETA: 4s - loss: 1.3468 - acc: 0.6847 - ETA: 4s - loss: 1.3386 - acc: 0.6863 - ETA: 4s - loss: 1.3305 - acc: 0.6867 - ETA: 4s - loss: 1.3185 - acc: 0.6885 - ETA: 3s - loss: 1.3086 - acc: 0.6905 - ETA: 3s - loss: 1.3009 - acc: 0.6922 - ETA: 3s - loss: 1.2963 - acc: 0.6932 - ETA: 3s - loss: 1.2907 - acc: 0.6932 - ETA: 3s - loss: 1.2831 - acc: 0.6951 - ETA: 3s - loss: 1.2799 - acc: 0.6962 - ETA: 3s - loss: 1.2765 - acc: 0.6974 - ETA: 3s - loss: 1.2711 - acc: 0.6983 - ETA: 3s - loss: 1.2639 - acc: 0.6998 - ETA: 3s - loss: 1.2579 - acc: 0.7002 - ETA: 3s - loss: 1.2573 - acc: 0.7002 - ETA: 3s - loss: 1.2507 - acc: 0.7015 - ETA: 3s - loss: 1.2453 - acc: 0.7025 - ETA: 3s - loss: 1.2400 - acc: 0.7035 - ETA: 3s - loss: 1.2360 - acc: 0.7041 - ETA: 2s - loss: 1.2305 - acc: 0.7053 - ETA: 2s - loss: 1.2230 - acc: 0.7071 - ETA: 2s - loss: 1.2167 - acc: 0.7084 - ETA: 2s - loss: 1.2109 - acc: 0.7096 - ETA: 2s - loss: 1.2048 - acc: 0.7109 - ETA: 2s - loss: 1.1964 - acc: 0.7127 - ETA: 2s - loss: 1.1944 - acc: 0.7128 - ETA: 2s - loss: 1.1881 - acc: 0.7143 - ETA: 2s - loss: 1.1828 - acc: 0.7151 - ETA: 2s - loss: 1.1801 - acc: 0.7150 - ETA: 2s - loss: 1.1754 - acc: 0.7158 - ETA: 2s - loss: 1.1713 - acc: 0.7165 - ETA: 2s - loss: 1.1655 - acc: 0.7177 - ETA: 2s - loss: 1.1582 - acc: 0.7186 - ETA: 2s - loss: 1.1550 - acc: 0.7192 - ETA: 1s - loss: 1.1514 - acc: 0.7200 - ETA: 1s - loss: 1.1474 - acc: 0.7208 - ETA: 1s - loss: 1.1460 - acc: 0.7206 - ETA: 1s - loss: 1.1425 - acc: 0.7210 - ETA: 1s - loss: 1.1366 - acc: 0.7224 - ETA: 1s - loss: 1.1349 - acc: 0.7225 - ETA: 1s - loss: 1.1294 - acc: 0.7233 - ETA: 1s - loss: 1.1258 - acc: 0.7240 - ETA: 1s - loss: 1.1239 - acc: 0.7245 - ETA: 1s - loss: 1.1210 - acc: 0.7253 - ETA: 1s - loss: 1.1209 - acc: 0.7255 - ETA: 1s - loss: 1.1174 - acc: 0.7257 - ETA: 1s - loss: 1.1125 - acc: 0.7270 - ETA: 1s - loss: 1.1077 - acc: 0.7275 - ETA: 1s - loss: 1.1030 - acc: 0.7285 - ETA: 1s - loss: 1.0988 - acc: 0.7293 - ETA: 0s - loss: 1.0965 - acc: 0.7297 - ETA: 0s - loss: 1.0921 - acc: 0.7304 - ETA: 0s - loss: 1.0877 - acc: 0.7311 - ETA: 0s - loss: 1.0833 - acc: 0.7321 - ETA: 0s - loss: 1.0787 - acc: 0.7330 - ETA: 0s - loss: 1.0759 - acc: 0.7339 - ETA: 0s - loss: 1.0699 - acc: 0.7350 - ETA: 0s - loss: 1.0669 - acc: 0.7354 - ETA: 0s - loss: 1.0652 - acc: 0.7358 - ETA: 0s - loss: 1.0638 - acc: 0.7357 - ETA: 0s - loss: 1.0611 - acc: 0.7359 - ETA: 0s - loss: 1.0581 - acc: 0.7363 - ETA: 0s - loss: 1.0537 - acc: 0.7373Epoch 00000: val_loss improved from inf to 0.51751, saving model to weights.best.Xception.hdf5
    6680/6680 [==============================] - 11s - loss: 1.0489 - acc: 0.7386 - val_loss: 0.5175 - val_acc: 0.8287
    Epoch 2/20
    6620/6680 [============================>.] - ETA: 7s - loss: 0.3979 - acc: 0.8500 - ETA: 7s - loss: 0.3478 - acc: 0.8500 - ETA: 7s - loss: 0.3562 - acc: 0.8714 - ETA: 7s - loss: 0.3885 - acc: 0.8650 - ETA: 7s - loss: 0.3679 - acc: 0.8692 - ETA: 7s - loss: 0.3545 - acc: 0.8719 - ETA: 7s - loss: 0.3575 - acc: 0.8737 - ETA: 6s - loss: 0.3531 - acc: 0.8750 - ETA: 6s - loss: 0.3631 - acc: 0.8720 - ETA: 6s - loss: 0.3674 - acc: 0.8750 - ETA: 6s - loss: 0.3743 - acc: 0.8742 - ETA: 6s - loss: 0.3755 - acc: 0.8750 - ETA: 6s - loss: 0.3889 - acc: 0.8757 - ETA: 6s - loss: 0.3845 - acc: 0.8787 - ETA: 6s - loss: 0.3848 - acc: 0.8791 - ETA: 6s - loss: 0.3822 - acc: 0.8783 - ETA: 6s - loss: 0.3915 - acc: 0.8765 - ETA: 6s - loss: 0.3991 - acc: 0.8712 - ETA: 6s - loss: 0.4072 - acc: 0.8673 - ETA: 6s - loss: 0.3966 - acc: 0.8724 - ETA: 6s - loss: 0.4024 - acc: 0.8689 - ETA: 5s - loss: 0.4074 - acc: 0.8672 - ETA: 5s - loss: 0.4160 - acc: 0.8634 - ETA: 5s - loss: 0.4107 - acc: 0.8657 - ETA: 5s - loss: 0.4102 - acc: 0.8651 - ETA: 5s - loss: 0.4076 - acc: 0.8651 - ETA: 5s - loss: 0.4050 - acc: 0.8658 - ETA: 5s - loss: 0.4067 - acc: 0.8652 - ETA: 5s - loss: 0.3986 - acc: 0.8682 - ETA: 5s - loss: 0.4027 - acc: 0.8665 - ETA: 5s - loss: 0.3969 - acc: 0.8681 - ETA: 5s - loss: 0.3979 - acc: 0.8681 - ETA: 5s - loss: 0.3987 - acc: 0.8675 - ETA: 5s - loss: 0.4006 - acc: 0.8675 - ETA: 5s - loss: 0.4008 - acc: 0.8675 - ETA: 5s - loss: 0.4022 - acc: 0.8665 - ETA: 4s - loss: 0.4065 - acc: 0.8665 - ETA: 4s - loss: 0.4075 - acc: 0.8674 - ETA: 4s - loss: 0.4108 - acc: 0.8665 - ETA: 4s - loss: 0.4084 - acc: 0.8682 - ETA: 4s - loss: 0.4085 - acc: 0.8690 - ETA: 4s - loss: 0.4053 - acc: 0.8698 - ETA: 4s - loss: 0.3998 - acc: 0.8717 - ETA: 4s - loss: 0.4016 - acc: 0.8708 - ETA: 4s - loss: 0.4003 - acc: 0.8722 - ETA: 4s - loss: 0.3987 - acc: 0.8721 - ETA: 4s - loss: 0.3959 - acc: 0.8730 - ETA: 4s - loss: 0.4032 - acc: 0.8718 - ETA: 4s - loss: 0.4025 - acc: 0.8731 - ETA: 4s - loss: 0.4017 - acc: 0.8736 - ETA: 4s - loss: 0.4002 - acc: 0.8748 - ETA: 3s - loss: 0.4011 - acc: 0.8744 - ETA: 3s - loss: 0.4079 - acc: 0.8729 - ETA: 3s - loss: 0.4070 - acc: 0.8725 - ETA: 3s - loss: 0.4030 - acc: 0.8736 - ETA: 3s - loss: 0.4059 - acc: 0.8735 - ETA: 3s - loss: 0.4022 - acc: 0.8749 - ETA: 3s - loss: 0.4008 - acc: 0.8747 - ETA: 3s - loss: 0.3980 - acc: 0.8749 - ETA: 3s - loss: 0.4005 - acc: 0.8739 - ETA: 3s - loss: 0.4033 - acc: 0.8735 - ETA: 3s - loss: 0.4025 - acc: 0.8734 - ETA: 3s - loss: 0.4028 - acc: 0.8738 - ETA: 3s - loss: 0.4013 - acc: 0.8739 - ETA: 3s - loss: 0.4014 - acc: 0.8744 - ETA: 3s - loss: 0.4039 - acc: 0.8737 - ETA: 2s - loss: 0.4063 - acc: 0.8736 - ETA: 2s - loss: 0.4062 - acc: 0.8738 - ETA: 2s - loss: 0.4063 - acc: 0.8741 - ETA: 2s - loss: 0.4065 - acc: 0.8743 - ETA: 2s - loss: 0.4064 - acc: 0.8737 - ETA: 2s - loss: 0.4060 - acc: 0.8745 - ETA: 2s - loss: 0.4054 - acc: 0.8744 - ETA: 2s - loss: 0.4108 - acc: 0.8727 - ETA: 2s - loss: 0.4100 - acc: 0.8733 - ETA: 2s - loss: 0.4084 - acc: 0.8741 - ETA: 2s - loss: 0.4072 - acc: 0.8747 - ETA: 2s - loss: 0.4082 - acc: 0.8746 - ETA: 2s - loss: 0.4084 - acc: 0.8740 - ETA: 2s - loss: 0.4073 - acc: 0.8748 - ETA: 2s - loss: 0.4061 - acc: 0.8747 - ETA: 1s - loss: 0.4082 - acc: 0.8746 - ETA: 1s - loss: 0.4086 - acc: 0.8743 - ETA: 1s - loss: 0.4104 - acc: 0.8736 - ETA: 1s - loss: 0.4101 - acc: 0.8737 - ETA: 1s - loss: 0.4070 - acc: 0.8746 - ETA: 1s - loss: 0.4047 - acc: 0.8753 - ETA: 1s - loss: 0.4035 - acc: 0.8754 - ETA: 1s - loss: 0.4038 - acc: 0.8755 - ETA: 1s - loss: 0.4050 - acc: 0.8743 - ETA: 1s - loss: 0.4059 - acc: 0.8740 - ETA: 1s - loss: 0.4074 - acc: 0.8735 - ETA: 1s - loss: 0.4077 - acc: 0.8736 - ETA: 1s - loss: 0.4063 - acc: 0.8743 - ETA: 1s - loss: 0.4072 - acc: 0.8742 - ETA: 1s - loss: 0.4055 - acc: 0.8747 - ETA: 0s - loss: 0.4073 - acc: 0.8744 - ETA: 0s - loss: 0.4058 - acc: 0.8748 - ETA: 0s - loss: 0.4029 - acc: 0.8758 - ETA: 0s - loss: 0.4026 - acc: 0.8753 - ETA: 0s - loss: 0.4013 - acc: 0.8754 - ETA: 0s - loss: 0.3989 - acc: 0.8762 - ETA: 0s - loss: 0.3989 - acc: 0.8759 - ETA: 0s - loss: 0.3960 - acc: 0.8769 - ETA: 0s - loss: 0.3959 - acc: 0.8768 - ETA: 0s - loss: 0.3962 - acc: 0.8764 - ETA: 0s - loss: 0.3972 - acc: 0.8765 - ETA: 0s - loss: 0.3974 - acc: 0.8761 - ETA: 0s - loss: 0.3981 - acc: 0.8760 - ETA: 0s - loss: 0.3991 - acc: 0.8758 - ETA: 0s - loss: 0.3967 - acc: 0.8766Epoch 00001: val_loss improved from 0.51751 to 0.49419, saving model to weights.best.Xception.hdf5
    6680/6680 [==============================] - 7s - loss: 0.3970 - acc: 0.8765 - val_loss: 0.4942 - val_acc: 0.8467
    Epoch 3/20
    6640/6680 [============================>.] - ETA: 6s - loss: 0.2117 - acc: 0.9500 - ETA: 6s - loss: 0.3529 - acc: 0.9125 - ETA: 6s - loss: 0.3027 - acc: 0.9000 - ETA: 6s - loss: 0.3290 - acc: 0.8950 - ETA: 6s - loss: 0.3522 - acc: 0.8885 - ETA: 6s - loss: 0.3138 - acc: 0.8969 - ETA: 6s - loss: 0.3289 - acc: 0.8974 - ETA: 6s - loss: 0.3111 - acc: 0.9023 - ETA: 6s - loss: 0.3087 - acc: 0.8960 - ETA: 6s - loss: 0.3035 - acc: 0.9000 - ETA: 6s - loss: 0.2889 - acc: 0.9048 - ETA: 6s - loss: 0.2840 - acc: 0.9088 - ETA: 6s - loss: 0.2783 - acc: 0.9095 - ETA: 6s - loss: 0.2810 - acc: 0.9087 - ETA: 6s - loss: 0.2698 - acc: 0.9128 - ETA: 6s - loss: 0.2898 - acc: 0.9065 - ETA: 6s - loss: 0.2967 - acc: 0.9051 - ETA: 6s - loss: 0.2896 - acc: 0.9067 - ETA: 5s - loss: 0.2954 - acc: 0.9045 - ETA: 5s - loss: 0.2979 - acc: 0.9052 - ETA: 5s - loss: 0.2950 - acc: 0.9074 - ETA: 5s - loss: 0.3010 - acc: 0.9086 - ETA: 5s - loss: 0.3164 - acc: 0.9037 - ETA: 5s - loss: 0.3164 - acc: 0.9043 - ETA: 5s - loss: 0.3249 - acc: 0.9027 - ETA: 5s - loss: 0.3204 - acc: 0.9033 - ETA: 5s - loss: 0.3175 - acc: 0.9032 - ETA: 5s - loss: 0.3323 - acc: 0.8988 - ETA: 5s - loss: 0.3274 - acc: 0.9000 - ETA: 5s - loss: 0.3212 - acc: 0.9017 - ETA: 5s - loss: 0.3189 - acc: 0.9016 - ETA: 5s - loss: 0.3228 - acc: 0.9000 - ETA: 5s - loss: 0.3217 - acc: 0.9005 - ETA: 5s - loss: 0.3211 - acc: 0.9015 - ETA: 4s - loss: 0.3208 - acc: 0.9019 - ETA: 4s - loss: 0.3238 - acc: 0.9005 - ETA: 4s - loss: 0.3209 - acc: 0.9014 - ETA: 4s - loss: 0.3240 - acc: 0.9009 - ETA: 4s - loss: 0.3246 - acc: 0.8996 - ETA: 4s - loss: 0.3243 - acc: 0.8996 - ETA: 4s - loss: 0.3244 - acc: 0.9004 - ETA: 4s - loss: 0.3211 - acc: 0.9016 - ETA: 4s - loss: 0.3182 - acc: 0.9024 - ETA: 4s - loss: 0.3195 - acc: 0.9019 - ETA: 4s - loss: 0.3174 - acc: 0.9019 - ETA: 4s - loss: 0.3152 - acc: 0.9022 - ETA: 4s - loss: 0.3161 - acc: 0.9022 - ETA: 4s - loss: 0.3182 - acc: 0.9028 - ETA: 4s - loss: 0.3187 - acc: 0.9017 - ETA: 4s - loss: 0.3183 - acc: 0.9024 - ETA: 3s - loss: 0.3141 - acc: 0.9037 - ETA: 3s - loss: 0.3120 - acc: 0.9046 - ETA: 3s - loss: 0.3084 - acc: 0.9058 - ETA: 3s - loss: 0.3059 - acc: 0.9063 - ETA: 3s - loss: 0.3081 - acc: 0.9052 - ETA: 3s - loss: 0.3085 - acc: 0.9045 - ETA: 3s - loss: 0.3092 - acc: 0.9045 - ETA: 3s - loss: 0.3124 - acc: 0.9035 - ETA: 3s - loss: 0.3151 - acc: 0.9032 - ETA: 3s - loss: 0.3136 - acc: 0.9034 - ETA: 3s - loss: 0.3147 - acc: 0.9033 - ETA: 3s - loss: 0.3166 - acc: 0.9036 - ETA: 3s - loss: 0.3173 - acc: 0.9032 - ETA: 3s - loss: 0.3161 - acc: 0.9032 - ETA: 3s - loss: 0.3191 - acc: 0.9018 - ETA: 2s - loss: 0.3205 - acc: 0.9015 - ETA: 2s - loss: 0.3223 - acc: 0.9015 - ETA: 2s - loss: 0.3191 - acc: 0.9027 - ETA: 2s - loss: 0.3212 - acc: 0.9027 - ETA: 2s - loss: 0.3245 - acc: 0.9014 - ETA: 2s - loss: 0.3275 - acc: 0.9002 - ETA: 2s - loss: 0.3277 - acc: 0.8998 - ETA: 2s - loss: 0.3308 - acc: 0.8991 - ETA: 2s - loss: 0.3283 - acc: 0.8995 - ETA: 2s - loss: 0.3290 - acc: 0.8991 - ETA: 2s - loss: 0.3277 - acc: 0.8993 - ETA: 2s - loss: 0.3260 - acc: 0.8998 - ETA: 2s - loss: 0.3253 - acc: 0.8994 - ETA: 2s - loss: 0.3238 - acc: 0.8998 - ETA: 2s - loss: 0.3236 - acc: 0.9000 - ETA: 2s - loss: 0.3244 - acc: 0.8996 - ETA: 1s - loss: 0.3254 - acc: 0.8981 - ETA: 1s - loss: 0.3250 - acc: 0.8980 - ETA: 1s - loss: 0.3265 - acc: 0.8972 - ETA: 1s - loss: 0.3280 - acc: 0.8972 - ETA: 1s - loss: 0.3287 - acc: 0.8967 - ETA: 1s - loss: 0.3274 - acc: 0.8971 - ETA: 1s - loss: 0.3265 - acc: 0.8977 - ETA: 1s - loss: 0.3260 - acc: 0.8979 - ETA: 1s - loss: 0.3251 - acc: 0.8981 - ETA: 1s - loss: 0.3229 - acc: 0.8987 - ETA: 1s - loss: 0.3214 - acc: 0.8991 - ETA: 1s - loss: 0.3217 - acc: 0.8989 - ETA: 1s - loss: 0.3212 - acc: 0.8987 - ETA: 1s - loss: 0.3232 - acc: 0.8982 - ETA: 1s - loss: 0.3216 - acc: 0.8984 - ETA: 0s - loss: 0.3191 - acc: 0.8993 - ETA: 0s - loss: 0.3200 - acc: 0.8993 - ETA: 0s - loss: 0.3210 - acc: 0.8991 - ETA: 0s - loss: 0.3224 - acc: 0.8987 - ETA: 0s - loss: 0.3209 - acc: 0.8992 - ETA: 0s - loss: 0.3225 - acc: 0.8988 - ETA: 0s - loss: 0.3248 - acc: 0.8989 - ETA: 0s - loss: 0.3239 - acc: 0.8990 - ETA: 0s - loss: 0.3236 - acc: 0.8992 - ETA: 0s - loss: 0.3232 - acc: 0.8995 - ETA: 0s - loss: 0.3214 - acc: 0.9002 - ETA: 0s - loss: 0.3219 - acc: 0.9005 - ETA: 0s - loss: 0.3239 - acc: 0.8998 - ETA: 0s - loss: 0.3249 - acc: 0.8995 - ETA: 0s - loss: 0.3240 - acc: 0.8997 - ETA: 0s - loss: 0.3232 - acc: 0.9000Epoch 00002: val_loss improved from 0.49419 to 0.48968, saving model to weights.best.Xception.hdf5
    6680/6680 [==============================] - 7s - loss: 0.3235 - acc: 0.8999 - val_loss: 0.4897 - val_acc: 0.8455
    Epoch 4/20
    6640/6680 [============================>.] - ETA: 6s - loss: 0.2434 - acc: 0.8500 - ETA: 6s - loss: 0.2229 - acc: 0.9000 - ETA: 6s - loss: 0.2553 - acc: 0.9071 - ETA: 6s - loss: 0.2461 - acc: 0.9100 - ETA: 6s - loss: 0.2156 - acc: 0.9231 - ETA: 6s - loss: 0.2247 - acc: 0.9187 - ETA: 6s - loss: 0.2415 - acc: 0.9105 - ETA: 6s - loss: 0.2547 - acc: 0.9114 - ETA: 6s - loss: 0.2454 - acc: 0.9180 - ETA: 6s - loss: 0.2304 - acc: 0.9214 - ETA: 6s - loss: 0.2393 - acc: 0.9145 - ETA: 6s - loss: 0.2477 - acc: 0.9147 - ETA: 6s - loss: 0.2553 - acc: 0.9135 - ETA: 6s - loss: 0.2492 - acc: 0.9162 - ETA: 6s - loss: 0.2475 - acc: 0.9174 - ETA: 6s - loss: 0.2441 - acc: 0.9196 - ETA: 6s - loss: 0.2489 - acc: 0.9194 - ETA: 6s - loss: 0.2410 - acc: 0.9221 - ETA: 5s - loss: 0.2414 - acc: 0.9218 - ETA: 5s - loss: 0.2383 - acc: 0.9207 - ETA: 5s - loss: 0.2422 - acc: 0.9180 - ETA: 5s - loss: 0.2422 - acc: 0.9180 - ETA: 5s - loss: 0.2392 - acc: 0.9194 - ETA: 5s - loss: 0.2348 - acc: 0.9207 - ETA: 5s - loss: 0.2367 - acc: 0.9199 - ETA: 5s - loss: 0.2391 - acc: 0.9197 - ETA: 5s - loss: 0.2416 - acc: 0.9196 - ETA: 5s - loss: 0.2449 - acc: 0.9195 - ETA: 5s - loss: 0.2495 - acc: 0.9176 - ETA: 5s - loss: 0.2430 - acc: 0.9205 - ETA: 5s - loss: 0.2404 - acc: 0.9203 - ETA: 5s - loss: 0.2412 - acc: 0.9210 - ETA: 5s - loss: 0.2425 - acc: 0.9198 - ETA: 5s - loss: 0.2452 - acc: 0.9187 - ETA: 5s - loss: 0.2435 - acc: 0.9191 - ETA: 4s - loss: 0.2430 - acc: 0.9190 - ETA: 4s - loss: 0.2434 - acc: 0.9190 - ETA: 4s - loss: 0.2427 - acc: 0.9194 - ETA: 4s - loss: 0.2431 - acc: 0.9197 - ETA: 4s - loss: 0.2405 - acc: 0.9209 - ETA: 4s - loss: 0.2395 - acc: 0.9221 - ETA: 4s - loss: 0.2396 - acc: 0.9220 - ETA: 4s - loss: 0.2384 - acc: 0.9222 - ETA: 4s - loss: 0.2361 - acc: 0.9236 - ETA: 4s - loss: 0.2375 - acc: 0.9223 - ETA: 4s - loss: 0.2368 - acc: 0.9226 - ETA: 4s - loss: 0.2382 - acc: 0.9225 - ETA: 4s - loss: 0.2404 - acc: 0.9223 - ETA: 4s - loss: 0.2402 - acc: 0.9222 - ETA: 4s - loss: 0.2412 - acc: 0.9218 - ETA: 3s - loss: 0.2459 - acc: 0.9210 - ETA: 3s - loss: 0.2496 - acc: 0.9199 - ETA: 3s - loss: 0.2568 - acc: 0.9183 - ETA: 3s - loss: 0.2542 - acc: 0.9192 - ETA: 3s - loss: 0.2585 - acc: 0.9176 - ETA: 3s - loss: 0.2587 - acc: 0.9176 - ETA: 3s - loss: 0.2577 - acc: 0.9182 - ETA: 3s - loss: 0.2570 - acc: 0.9184 - ETA: 3s - loss: 0.2584 - acc: 0.9172 - ETA: 3s - loss: 0.2579 - acc: 0.9169 - ETA: 3s - loss: 0.2562 - acc: 0.9175 - ETA: 3s - loss: 0.2573 - acc: 0.9175 - ETA: 3s - loss: 0.2562 - acc: 0.9177 - ETA: 3s - loss: 0.2560 - acc: 0.9175 - ETA: 3s - loss: 0.2544 - acc: 0.9180 - ETA: 2s - loss: 0.2543 - acc: 0.9182 - ETA: 2s - loss: 0.2546 - acc: 0.9184 - ETA: 2s - loss: 0.2533 - acc: 0.9184 - ETA: 2s - loss: 0.2549 - acc: 0.9179 - ETA: 2s - loss: 0.2552 - acc: 0.9181 - ETA: 2s - loss: 0.2580 - acc: 0.9174 - ETA: 2s - loss: 0.2575 - acc: 0.9171 - ETA: 2s - loss: 0.2584 - acc: 0.9167 - ETA: 2s - loss: 0.2579 - acc: 0.9169 - ETA: 2s - loss: 0.2596 - acc: 0.9167 - ETA: 2s - loss: 0.2593 - acc: 0.9164 - ETA: 2s - loss: 0.2580 - acc: 0.9171 - ETA: 2s - loss: 0.2602 - acc: 0.9169 - ETA: 2s - loss: 0.2636 - acc: 0.9162 - ETA: 2s - loss: 0.2638 - acc: 0.9160 - ETA: 2s - loss: 0.2617 - acc: 0.9167 - ETA: 1s - loss: 0.2621 - acc: 0.9167 - ETA: 1s - loss: 0.2663 - acc: 0.9154 - ETA: 1s - loss: 0.2714 - acc: 0.9151 - ETA: 1s - loss: 0.2731 - acc: 0.9145 - ETA: 1s - loss: 0.2749 - acc: 0.9147 - ETA: 1s - loss: 0.2756 - acc: 0.9141 - ETA: 1s - loss: 0.2773 - acc: 0.9132 - ETA: 1s - loss: 0.2761 - acc: 0.9134 - ETA: 1s - loss: 0.2749 - acc: 0.9142 - ETA: 1s - loss: 0.2754 - acc: 0.9139 - ETA: 1s - loss: 0.2755 - acc: 0.9137 - ETA: 1s - loss: 0.2750 - acc: 0.9141 - ETA: 1s - loss: 0.2753 - acc: 0.9138 - ETA: 1s - loss: 0.2778 - acc: 0.9137 - ETA: 1s - loss: 0.2781 - acc: 0.9137 - ETA: 1s - loss: 0.2781 - acc: 0.9134 - ETA: 0s - loss: 0.2801 - acc: 0.9133 - ETA: 0s - loss: 0.2783 - acc: 0.9137 - ETA: 0s - loss: 0.2772 - acc: 0.9139 - ETA: 0s - loss: 0.2766 - acc: 0.9142 - ETA: 0s - loss: 0.2768 - acc: 0.9141 - ETA: 0s - loss: 0.2771 - acc: 0.9138 - ETA: 0s - loss: 0.2769 - acc: 0.9138 - ETA: 0s - loss: 0.2767 - acc: 0.9135 - ETA: 0s - loss: 0.2769 - acc: 0.9137 - ETA: 0s - loss: 0.2772 - acc: 0.9134 - ETA: 0s - loss: 0.2768 - acc: 0.9134 - ETA: 0s - loss: 0.2771 - acc: 0.9132 - ETA: 0s - loss: 0.2814 - acc: 0.9124 - ETA: 0s - loss: 0.2821 - acc: 0.9122 - ETA: 0s - loss: 0.2809 - acc: 0.9125Epoch 00003: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.2805 - acc: 0.9124 - val_loss: 0.4960 - val_acc: 0.8551
    Epoch 5/20
    6660/6680 [============================>.] - ETA: 6s - loss: 0.2675 - acc: 0.9500 - ETA: 6s - loss: 0.2163 - acc: 0.9500 - ETA: 6s - loss: 0.1891 - acc: 0.9429 - ETA: 6s - loss: 0.1682 - acc: 0.9500 - ETA: 6s - loss: 0.1567 - acc: 0.9462 - ETA: 6s - loss: 0.1586 - acc: 0.9437 - ETA: 6s - loss: 0.1491 - acc: 0.9500 - ETA: 6s - loss: 0.1609 - acc: 0.9477 - ETA: 6s - loss: 0.1563 - acc: 0.9520 - ETA: 6s - loss: 0.1601 - acc: 0.9482 - ETA: 6s - loss: 0.1729 - acc: 0.9468 - ETA: 6s - loss: 0.1730 - acc: 0.9471 - ETA: 6s - loss: 0.1810 - acc: 0.9459 - ETA: 6s - loss: 0.1916 - acc: 0.9400 - ETA: 6s - loss: 0.1947 - acc: 0.9407 - ETA: 6s - loss: 0.1905 - acc: 0.9424 - ETA: 6s - loss: 0.2041 - acc: 0.9357 - ETA: 5s - loss: 0.2072 - acc: 0.9365 - ETA: 5s - loss: 0.2120 - acc: 0.9364 - ETA: 5s - loss: 0.2120 - acc: 0.9353 - ETA: 5s - loss: 0.2109 - acc: 0.9352 - ETA: 5s - loss: 0.2087 - acc: 0.9344 - ETA: 5s - loss: 0.2130 - acc: 0.9321 - ETA: 5s - loss: 0.2103 - acc: 0.9314 - ETA: 5s - loss: 0.2135 - acc: 0.9301 - ETA: 5s - loss: 0.2163 - acc: 0.9289 - ETA: 5s - loss: 0.2142 - acc: 0.9297 - ETA: 5s - loss: 0.2115 - acc: 0.9299 - ETA: 5s - loss: 0.2098 - acc: 0.9300 - ETA: 5s - loss: 0.2107 - acc: 0.9295 - ETA: 5s - loss: 0.2101 - acc: 0.9297 - ETA: 5s - loss: 0.2078 - acc: 0.9309 - ETA: 5s - loss: 0.2032 - acc: 0.9325 - ETA: 4s - loss: 0.2035 - acc: 0.9330 - ETA: 4s - loss: 0.2044 - acc: 0.9325 - ETA: 4s - loss: 0.2051 - acc: 0.9325 - ETA: 4s - loss: 0.2015 - acc: 0.9339 - ETA: 4s - loss: 0.2017 - acc: 0.9339 - ETA: 4s - loss: 0.2015 - acc: 0.9330 - ETA: 4s - loss: 0.1997 - acc: 0.9339 - ETA: 4s - loss: 0.1976 - acc: 0.9343 - ETA: 4s - loss: 0.1962 - acc: 0.9347 - ETA: 4s - loss: 0.2039 - acc: 0.9339 - ETA: 4s - loss: 0.2081 - acc: 0.9331 - ETA: 4s - loss: 0.2054 - acc: 0.9338 - ETA: 4s - loss: 0.2021 - acc: 0.9349 - ETA: 4s - loss: 0.1992 - acc: 0.9360 - ETA: 4s - loss: 0.2056 - acc: 0.9352 - ETA: 4s - loss: 0.2031 - acc: 0.9359 - ETA: 3s - loss: 0.2063 - acc: 0.9348 - ETA: 3s - loss: 0.2036 - acc: 0.9361 - ETA: 3s - loss: 0.2090 - acc: 0.9347 - ETA: 3s - loss: 0.2117 - acc: 0.9338 - ETA: 3s - loss: 0.2136 - acc: 0.9334 - ETA: 3s - loss: 0.2140 - acc: 0.9334 - ETA: 3s - loss: 0.2127 - acc: 0.9337 - ETA: 3s - loss: 0.2150 - acc: 0.9325 - ETA: 3s - loss: 0.2169 - acc: 0.9314 - ETA: 3s - loss: 0.2176 - acc: 0.9309 - ETA: 3s - loss: 0.2182 - acc: 0.9309 - ETA: 3s - loss: 0.2173 - acc: 0.9312 - ETA: 3s - loss: 0.2176 - acc: 0.9310 - ETA: 3s - loss: 0.2150 - acc: 0.9321 - ETA: 3s - loss: 0.2149 - acc: 0.9321 - ETA: 3s - loss: 0.2201 - acc: 0.9321 - ETA: 2s - loss: 0.2193 - acc: 0.9321 - ETA: 2s - loss: 0.2232 - acc: 0.9314 - ETA: 2s - loss: 0.2226 - acc: 0.9314 - ETA: 2s - loss: 0.2262 - acc: 0.9300 - ETA: 2s - loss: 0.2257 - acc: 0.9300 - ETA: 2s - loss: 0.2251 - acc: 0.9299 - ETA: 2s - loss: 0.2244 - acc: 0.9299 - ETA: 2s - loss: 0.2273 - acc: 0.9283 - ETA: 2s - loss: 0.2284 - acc: 0.9275 - ETA: 2s - loss: 0.2275 - acc: 0.9275 - ETA: 2s - loss: 0.2310 - acc: 0.9278 - ETA: 2s - loss: 0.2306 - acc: 0.9281 - ETA: 2s - loss: 0.2349 - acc: 0.9268 - ETA: 2s - loss: 0.2358 - acc: 0.9261 - ETA: 2s - loss: 0.2369 - acc: 0.9253 - ETA: 2s - loss: 0.2375 - acc: 0.9252 - ETA: 1s - loss: 0.2369 - acc: 0.9255 - ETA: 1s - loss: 0.2370 - acc: 0.9256 - ETA: 1s - loss: 0.2373 - acc: 0.9259 - ETA: 1s - loss: 0.2354 - acc: 0.9266 - ETA: 1s - loss: 0.2383 - acc: 0.9259 - ETA: 1s - loss: 0.2388 - acc: 0.9258 - ETA: 1s - loss: 0.2404 - acc: 0.9255 - ETA: 1s - loss: 0.2409 - acc: 0.9256 - ETA: 1s - loss: 0.2407 - acc: 0.9258 - ETA: 1s - loss: 0.2405 - acc: 0.9257 - ETA: 1s - loss: 0.2409 - acc: 0.9253 - ETA: 1s - loss: 0.2398 - acc: 0.9255 - ETA: 1s - loss: 0.2390 - acc: 0.9258 - ETA: 1s - loss: 0.2396 - acc: 0.9252 - ETA: 1s - loss: 0.2429 - acc: 0.9240 - ETA: 0s - loss: 0.2436 - acc: 0.9238 - ETA: 0s - loss: 0.2456 - acc: 0.9234 - ETA: 0s - loss: 0.2446 - acc: 0.9238 - ETA: 0s - loss: 0.2442 - acc: 0.9239 - ETA: 0s - loss: 0.2435 - acc: 0.9243 - ETA: 0s - loss: 0.2424 - acc: 0.9248 - ETA: 0s - loss: 0.2451 - acc: 0.9243 - ETA: 0s - loss: 0.2449 - acc: 0.9246 - ETA: 0s - loss: 0.2460 - acc: 0.9245 - ETA: 0s - loss: 0.2463 - acc: 0.9244 - ETA: 0s - loss: 0.2460 - acc: 0.9245 - ETA: 0s - loss: 0.2461 - acc: 0.9241 - ETA: 0s - loss: 0.2451 - acc: 0.9242 - ETA: 0s - loss: 0.2447 - acc: 0.9243 - ETA: 0s - loss: 0.2433 - acc: 0.9245 - ETA: 0s - loss: 0.2444 - acc: 0.9242Epoch 00004: val_loss improved from 0.48968 to 0.48411, saving model to weights.best.Xception.hdf5
    6680/6680 [==============================] - 7s - loss: 0.2448 - acc: 0.9241 - val_loss: 0.4841 - val_acc: 0.8587
    Epoch 6/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.2373 - acc: 0.9000 - ETA: 6s - loss: 0.1525 - acc: 0.9500 - ETA: 6s - loss: 0.1618 - acc: 0.9429 - ETA: 6s - loss: 0.1618 - acc: 0.9500 - ETA: 6s - loss: 0.1528 - acc: 0.9538 - ETA: 6s - loss: 0.1455 - acc: 0.9500 - ETA: 6s - loss: 0.1410 - acc: 0.9500 - ETA: 6s - loss: 0.1423 - acc: 0.9523 - ETA: 6s - loss: 0.1553 - acc: 0.9480 - ETA: 6s - loss: 0.1771 - acc: 0.9446 - ETA: 6s - loss: 0.1784 - acc: 0.9419 - ETA: 6s - loss: 0.1705 - acc: 0.9441 - ETA: 6s - loss: 0.1732 - acc: 0.9432 - ETA: 6s - loss: 0.1697 - acc: 0.9437 - ETA: 6s - loss: 0.1740 - acc: 0.9419 - ETA: 6s - loss: 0.1750 - acc: 0.9391 - ETA: 6s - loss: 0.1713 - acc: 0.9408 - ETA: 6s - loss: 0.1653 - acc: 0.9423 - ETA: 6s - loss: 0.1715 - acc: 0.9400 - ETA: 5s - loss: 0.1749 - acc: 0.9397 - ETA: 5s - loss: 0.1718 - acc: 0.9402 - ETA: 5s - loss: 0.1775 - acc: 0.9406 - ETA: 5s - loss: 0.1760 - acc: 0.9410 - ETA: 5s - loss: 0.1788 - acc: 0.9393 - ETA: 5s - loss: 0.1767 - acc: 0.9397 - ETA: 5s - loss: 0.1743 - acc: 0.9408 - ETA: 5s - loss: 0.1750 - acc: 0.9405 - ETA: 5s - loss: 0.1717 - acc: 0.9415 - ETA: 5s - loss: 0.1726 - acc: 0.9406 - ETA: 5s - loss: 0.1742 - acc: 0.9398 - ETA: 5s - loss: 0.1725 - acc: 0.9396 - ETA: 5s - loss: 0.1771 - acc: 0.9383 - ETA: 5s - loss: 0.1797 - acc: 0.9381 - ETA: 4s - loss: 0.1782 - acc: 0.9395 - ETA: 4s - loss: 0.1777 - acc: 0.9393 - ETA: 4s - loss: 0.1756 - acc: 0.9401 - ETA: 4s - loss: 0.1766 - acc: 0.9408 - ETA: 4s - loss: 0.1770 - acc: 0.9402 - ETA: 4s - loss: 0.1788 - acc: 0.9409 - ETA: 4s - loss: 0.1787 - acc: 0.9407 - ETA: 4s - loss: 0.1839 - acc: 0.9384 - ETA: 4s - loss: 0.1910 - acc: 0.9379 - ETA: 4s - loss: 0.1882 - acc: 0.9386 - ETA: 4s - loss: 0.1940 - acc: 0.9377 - ETA: 4s - loss: 0.1944 - acc: 0.9380 - ETA: 4s - loss: 0.1934 - acc: 0.9379 - ETA: 4s - loss: 0.1944 - acc: 0.9378 - ETA: 4s - loss: 0.1934 - acc: 0.9380 - ETA: 4s - loss: 0.1937 - acc: 0.9372 - ETA: 3s - loss: 0.2007 - acc: 0.9361 - ETA: 3s - loss: 0.1987 - acc: 0.9368 - ETA: 3s - loss: 0.1980 - acc: 0.9370 - ETA: 3s - loss: 0.1966 - acc: 0.9369 - ETA: 3s - loss: 0.1982 - acc: 0.9366 - ETA: 3s - loss: 0.1982 - acc: 0.9365 - ETA: 3s - loss: 0.1972 - acc: 0.9370 - ETA: 3s - loss: 0.1961 - acc: 0.9370 - ETA: 3s - loss: 0.1934 - acc: 0.9381 - ETA: 3s - loss: 0.1942 - acc: 0.9377 - ETA: 3s - loss: 0.1955 - acc: 0.9374 - ETA: 3s - loss: 0.1956 - acc: 0.9370 - ETA: 3s - loss: 0.1951 - acc: 0.9372 - ETA: 3s - loss: 0.1943 - acc: 0.9374 - ETA: 3s - loss: 0.1927 - acc: 0.9376 - ETA: 3s - loss: 0.1943 - acc: 0.9376 - ETA: 2s - loss: 0.1982 - acc: 0.9370 - ETA: 2s - loss: 0.1975 - acc: 0.9372 - ETA: 2s - loss: 0.1978 - acc: 0.9369 - ETA: 2s - loss: 0.1989 - acc: 0.9366 - ETA: 2s - loss: 0.2019 - acc: 0.9361 - ETA: 2s - loss: 0.2081 - acc: 0.9348 - ETA: 2s - loss: 0.2066 - acc: 0.9353 - ETA: 2s - loss: 0.2064 - acc: 0.9350 - ETA: 2s - loss: 0.2062 - acc: 0.9352 - ETA: 2s - loss: 0.2044 - acc: 0.9359 - ETA: 2s - loss: 0.2054 - acc: 0.9356 - ETA: 2s - loss: 0.2060 - acc: 0.9349 - ETA: 2s - loss: 0.2051 - acc: 0.9351 - ETA: 2s - loss: 0.2071 - acc: 0.9349 - ETA: 2s - loss: 0.2070 - acc: 0.9349 - ETA: 1s - loss: 0.2100 - acc: 0.9346 - ETA: 1s - loss: 0.2096 - acc: 0.9348 - ETA: 1s - loss: 0.2082 - acc: 0.9350 - ETA: 1s - loss: 0.2069 - acc: 0.9356 - ETA: 1s - loss: 0.2080 - acc: 0.9354 - ETA: 1s - loss: 0.2077 - acc: 0.9354 - ETA: 1s - loss: 0.2085 - acc: 0.9359 - ETA: 1s - loss: 0.2132 - acc: 0.9349 - ETA: 1s - loss: 0.2134 - acc: 0.9349 - ETA: 1s - loss: 0.2132 - acc: 0.9351 - ETA: 1s - loss: 0.2121 - acc: 0.9352 - ETA: 1s - loss: 0.2162 - acc: 0.9347 - ETA: 1s - loss: 0.2171 - acc: 0.9345 - ETA: 1s - loss: 0.2177 - acc: 0.9345 - ETA: 1s - loss: 0.2177 - acc: 0.9341 - ETA: 1s - loss: 0.2192 - acc: 0.9336 - ETA: 0s - loss: 0.2191 - acc: 0.9337 - ETA: 0s - loss: 0.2197 - acc: 0.9334 - ETA: 0s - loss: 0.2188 - acc: 0.9334 - ETA: 0s - loss: 0.2183 - acc: 0.9334 - ETA: 0s - loss: 0.2182 - acc: 0.9337 - ETA: 0s - loss: 0.2178 - acc: 0.9339 - ETA: 0s - loss: 0.2181 - acc: 0.9337 - ETA: 0s - loss: 0.2210 - acc: 0.9327 - ETA: 0s - loss: 0.2207 - acc: 0.9331 - ETA: 0s - loss: 0.2203 - acc: 0.9329 - ETA: 0s - loss: 0.2205 - acc: 0.9328 - ETA: 0s - loss: 0.2201 - acc: 0.9326 - ETA: 0s - loss: 0.2224 - acc: 0.9317 - ETA: 0s - loss: 0.2236 - acc: 0.9314 - ETA: 0s - loss: 0.2232 - acc: 0.9316Epoch 00005: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.2238 - acc: 0.9314 - val_loss: 0.5198 - val_acc: 0.8647
    Epoch 7/20
    6620/6680 [============================>.] - ETA: 7s - loss: 0.0120 - acc: 1.0000 - ETA: 6s - loss: 0.0314 - acc: 0.9875 - ETA: 6s - loss: 0.0841 - acc: 0.9714 - ETA: 6s - loss: 0.1783 - acc: 0.9550 - ETA: 6s - loss: 0.1598 - acc: 0.9615 - ETA: 6s - loss: 0.1698 - acc: 0.9562 - ETA: 6s - loss: 0.1943 - acc: 0.9447 - ETA: 6s - loss: 0.1962 - acc: 0.9341 - ETA: 6s - loss: 0.1907 - acc: 0.9340 - ETA: 6s - loss: 0.1915 - acc: 0.9357 - ETA: 6s - loss: 0.1866 - acc: 0.9355 - ETA: 6s - loss: 0.1879 - acc: 0.9324 - ETA: 6s - loss: 0.1906 - acc: 0.9311 - ETA: 6s - loss: 0.2003 - acc: 0.9300 - ETA: 6s - loss: 0.2009 - acc: 0.9291 - ETA: 6s - loss: 0.1899 - acc: 0.9337 - ETA: 6s - loss: 0.1847 - acc: 0.9357 - ETA: 6s - loss: 0.1816 - acc: 0.9375 - ETA: 5s - loss: 0.1840 - acc: 0.9355 - ETA: 5s - loss: 0.1798 - acc: 0.9362 - ETA: 5s - loss: 0.1827 - acc: 0.9336 - ETA: 5s - loss: 0.1840 - acc: 0.9320 - ETA: 5s - loss: 0.1867 - acc: 0.9321 - ETA: 5s - loss: 0.1842 - acc: 0.9336 - ETA: 5s - loss: 0.1915 - acc: 0.9336 - ETA: 5s - loss: 0.1889 - acc: 0.9342 - ETA: 5s - loss: 0.1874 - acc: 0.9348 - ETA: 5s - loss: 0.1860 - acc: 0.9360 - ETA: 5s - loss: 0.1806 - acc: 0.9382 - ETA: 5s - loss: 0.1816 - acc: 0.9381 - ETA: 5s - loss: 0.1846 - acc: 0.9374 - ETA: 5s - loss: 0.1895 - acc: 0.9362 - ETA: 5s - loss: 0.1893 - acc: 0.9366 - ETA: 4s - loss: 0.1889 - acc: 0.9370 - ETA: 4s - loss: 0.1865 - acc: 0.9374 - ETA: 4s - loss: 0.1893 - acc: 0.9373 - ETA: 4s - loss: 0.1848 - acc: 0.9390 - ETA: 4s - loss: 0.1824 - acc: 0.9393 - ETA: 4s - loss: 0.1851 - acc: 0.9383 - ETA: 4s - loss: 0.1884 - acc: 0.9373 - ETA: 4s - loss: 0.1953 - acc: 0.9364 - ETA: 4s - loss: 0.1932 - acc: 0.9367 - ETA: 4s - loss: 0.1901 - acc: 0.9378 - ETA: 4s - loss: 0.1895 - acc: 0.9381 - ETA: 4s - loss: 0.1881 - acc: 0.9380 - ETA: 4s - loss: 0.1870 - acc: 0.9375 - ETA: 4s - loss: 0.1894 - acc: 0.9367 - ETA: 4s - loss: 0.1893 - acc: 0.9359 - ETA: 4s - loss: 0.1910 - acc: 0.9362 - ETA: 3s - loss: 0.1898 - acc: 0.9365 - ETA: 3s - loss: 0.1914 - acc: 0.9358 - ETA: 3s - loss: 0.1898 - acc: 0.9360 - ETA: 3s - loss: 0.1877 - acc: 0.9366 - ETA: 3s - loss: 0.1899 - acc: 0.9359 - ETA: 3s - loss: 0.1898 - acc: 0.9359 - ETA: 3s - loss: 0.1886 - acc: 0.9367 - ETA: 3s - loss: 0.1878 - acc: 0.9373 - ETA: 3s - loss: 0.1859 - acc: 0.9381 - ETA: 3s - loss: 0.1858 - acc: 0.9377 - ETA: 3s - loss: 0.1864 - acc: 0.9376 - ETA: 3s - loss: 0.1866 - acc: 0.9381 - ETA: 3s - loss: 0.1854 - acc: 0.9386 - ETA: 3s - loss: 0.1866 - acc: 0.9385 - ETA: 3s - loss: 0.1892 - acc: 0.9371 - ETA: 3s - loss: 0.1892 - acc: 0.9368 - ETA: 2s - loss: 0.1903 - acc: 0.9365 - ETA: 2s - loss: 0.1902 - acc: 0.9362 - ETA: 2s - loss: 0.1911 - acc: 0.9361 - ETA: 2s - loss: 0.1930 - acc: 0.9366 - ETA: 2s - loss: 0.1914 - acc: 0.9370 - ETA: 2s - loss: 0.1926 - acc: 0.9370 - ETA: 2s - loss: 0.1941 - acc: 0.9369 - ETA: 2s - loss: 0.1946 - acc: 0.9366 - ETA: 2s - loss: 0.1964 - acc: 0.9364 - ETA: 2s - loss: 0.1976 - acc: 0.9361 - ETA: 2s - loss: 0.1995 - acc: 0.9356 - ETA: 2s - loss: 0.2007 - acc: 0.9356 - ETA: 2s - loss: 0.2004 - acc: 0.9353 - ETA: 2s - loss: 0.2005 - acc: 0.9355 - ETA: 2s - loss: 0.1998 - acc: 0.9357 - ETA: 1s - loss: 0.1991 - acc: 0.9357 - ETA: 1s - loss: 0.1976 - acc: 0.9361 - ETA: 1s - loss: 0.1964 - acc: 0.9362 - ETA: 1s - loss: 0.1975 - acc: 0.9360 - ETA: 1s - loss: 0.1970 - acc: 0.9364 - ETA: 1s - loss: 0.1958 - acc: 0.9365 - ETA: 1s - loss: 0.1970 - acc: 0.9367 - ETA: 1s - loss: 0.1979 - acc: 0.9366 - ETA: 1s - loss: 0.1979 - acc: 0.9368 - ETA: 1s - loss: 0.1972 - acc: 0.9371 - ETA: 1s - loss: 0.1978 - acc: 0.9371 - ETA: 1s - loss: 0.1983 - acc: 0.9365 - ETA: 1s - loss: 0.1994 - acc: 0.9363 - ETA: 1s - loss: 0.1994 - acc: 0.9364 - ETA: 1s - loss: 0.1998 - acc: 0.9364 - ETA: 1s - loss: 0.1994 - acc: 0.9364 - ETA: 0s - loss: 0.2005 - acc: 0.9356 - ETA: 0s - loss: 0.1993 - acc: 0.9358 - ETA: 0s - loss: 0.1989 - acc: 0.9359 - ETA: 0s - loss: 0.1998 - acc: 0.9361 - ETA: 0s - loss: 0.1993 - acc: 0.9359 - ETA: 0s - loss: 0.1990 - acc: 0.9360 - ETA: 0s - loss: 0.1995 - acc: 0.9357 - ETA: 0s - loss: 0.1985 - acc: 0.9360 - ETA: 0s - loss: 0.1988 - acc: 0.9358 - ETA: 0s - loss: 0.1974 - acc: 0.9362 - ETA: 0s - loss: 0.1982 - acc: 0.9362 - ETA: 0s - loss: 0.1984 - acc: 0.9365 - ETA: 0s - loss: 0.1986 - acc: 0.9363 - ETA: 0s - loss: 0.1980 - acc: 0.9366 - ETA: 0s - loss: 0.1985 - acc: 0.9364Epoch 00006: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1981 - acc: 0.9364 - val_loss: 0.5117 - val_acc: 0.8539
    Epoch 8/20
    6660/6680 [============================>.] - ETA: 6s - loss: 0.0584 - acc: 1.0000 - ETA: 6s - loss: 0.1994 - acc: 0.9500 - ETA: 6s - loss: 0.1597 - acc: 0.9571 - ETA: 6s - loss: 0.1994 - acc: 0.9500 - ETA: 6s - loss: 0.2007 - acc: 0.9538 - ETA: 6s - loss: 0.1783 - acc: 0.9594 - ETA: 6s - loss: 0.1587 - acc: 0.9632 - ETA: 6s - loss: 0.1664 - acc: 0.9591 - ETA: 6s - loss: 0.1552 - acc: 0.9600 - ETA: 6s - loss: 0.1599 - acc: 0.9589 - ETA: 6s - loss: 0.1725 - acc: 0.9565 - ETA: 6s - loss: 0.1667 - acc: 0.9574 - ETA: 6s - loss: 0.1592 - acc: 0.9595 - ETA: 6s - loss: 0.1551 - acc: 0.9587 - ETA: 6s - loss: 0.1575 - acc: 0.9593 - ETA: 6s - loss: 0.1673 - acc: 0.9565 - ETA: 6s - loss: 0.1596 - acc: 0.9592 - ETA: 6s - loss: 0.1569 - acc: 0.9596 - ETA: 5s - loss: 0.1621 - acc: 0.9573 - ETA: 5s - loss: 0.1584 - acc: 0.9586 - ETA: 5s - loss: 0.1562 - acc: 0.9574 - ETA: 5s - loss: 0.1509 - acc: 0.9586 - ETA: 5s - loss: 0.1468 - acc: 0.9597 - ETA: 5s - loss: 0.1476 - acc: 0.9600 - ETA: 5s - loss: 0.1443 - acc: 0.9610 - ETA: 5s - loss: 0.1442 - acc: 0.9599 - ETA: 5s - loss: 0.1415 - acc: 0.9601 - ETA: 5s - loss: 0.1437 - acc: 0.9591 - ETA: 5s - loss: 0.1419 - acc: 0.9594 - ETA: 5s - loss: 0.1440 - acc: 0.9580 - ETA: 5s - loss: 0.1427 - acc: 0.9577 - ETA: 5s - loss: 0.1430 - acc: 0.9574 - ETA: 4s - loss: 0.1422 - acc: 0.9577 - ETA: 4s - loss: 0.1479 - acc: 0.9565 - ETA: 4s - loss: 0.1458 - acc: 0.9568 - ETA: 4s - loss: 0.1460 - acc: 0.9561 - ETA: 4s - loss: 0.1444 - acc: 0.9564 - ETA: 4s - loss: 0.1420 - acc: 0.9567 - ETA: 4s - loss: 0.1422 - acc: 0.9574 - ETA: 4s - loss: 0.1423 - acc: 0.9564 - ETA: 4s - loss: 0.1440 - acc: 0.9554 - ETA: 4s - loss: 0.1420 - acc: 0.9560 - ETA: 4s - loss: 0.1432 - acc: 0.9563 - ETA: 4s - loss: 0.1432 - acc: 0.9565 - ETA: 4s - loss: 0.1467 - acc: 0.9553 - ETA: 4s - loss: 0.1466 - acc: 0.9551 - ETA: 4s - loss: 0.1498 - acc: 0.9550 - ETA: 4s - loss: 0.1577 - acc: 0.9535 - ETA: 3s - loss: 0.1553 - acc: 0.9541 - ETA: 3s - loss: 0.1551 - acc: 0.9541 - ETA: 3s - loss: 0.1543 - acc: 0.9543 - ETA: 3s - loss: 0.1523 - acc: 0.9549 - ETA: 3s - loss: 0.1527 - acc: 0.9545 - ETA: 3s - loss: 0.1533 - acc: 0.9537 - ETA: 3s - loss: 0.1574 - acc: 0.9540 - ETA: 3s - loss: 0.1581 - acc: 0.9539 - ETA: 3s - loss: 0.1577 - acc: 0.9541 - ETA: 3s - loss: 0.1615 - acc: 0.9529 - ETA: 3s - loss: 0.1609 - acc: 0.9526 - ETA: 3s - loss: 0.1604 - acc: 0.9528 - ETA: 3s - loss: 0.1621 - acc: 0.9519 - ETA: 3s - loss: 0.1625 - acc: 0.9516 - ETA: 3s - loss: 0.1639 - acc: 0.9513 - ETA: 3s - loss: 0.1646 - acc: 0.9513 - ETA: 2s - loss: 0.1630 - acc: 0.9518 - ETA: 2s - loss: 0.1656 - acc: 0.9500 - ETA: 2s - loss: 0.1642 - acc: 0.9505 - ETA: 2s - loss: 0.1622 - acc: 0.9512 - ETA: 2s - loss: 0.1667 - acc: 0.9502 - ETA: 2s - loss: 0.1681 - acc: 0.9495 - ETA: 2s - loss: 0.1679 - acc: 0.9498 - ETA: 2s - loss: 0.1675 - acc: 0.9498 - ETA: 2s - loss: 0.1674 - acc: 0.9495 - ETA: 2s - loss: 0.1706 - acc: 0.9493 - ETA: 2s - loss: 0.1716 - acc: 0.9484 - ETA: 2s - loss: 0.1701 - acc: 0.9489 - ETA: 2s - loss: 0.1693 - acc: 0.9491 - ETA: 2s - loss: 0.1707 - acc: 0.9485 - ETA: 2s - loss: 0.1708 - acc: 0.9485 - ETA: 2s - loss: 0.1692 - acc: 0.9489 - ETA: 1s - loss: 0.1695 - acc: 0.9485 - ETA: 1s - loss: 0.1688 - acc: 0.9488 - ETA: 1s - loss: 0.1685 - acc: 0.9490 - ETA: 1s - loss: 0.1699 - acc: 0.9488 - ETA: 1s - loss: 0.1692 - acc: 0.9490 - ETA: 1s - loss: 0.1691 - acc: 0.9490 - ETA: 1s - loss: 0.1698 - acc: 0.9486 - ETA: 1s - loss: 0.1700 - acc: 0.9485 - ETA: 1s - loss: 0.1711 - acc: 0.9483 - ETA: 1s - loss: 0.1707 - acc: 0.9487 - ETA: 1s - loss: 0.1729 - acc: 0.9483 - ETA: 1s - loss: 0.1741 - acc: 0.9485 - ETA: 1s - loss: 0.1733 - acc: 0.9487 - ETA: 1s - loss: 0.1728 - acc: 0.9487 - ETA: 1s - loss: 0.1725 - acc: 0.9488 - ETA: 1s - loss: 0.1730 - acc: 0.9484 - ETA: 0s - loss: 0.1730 - acc: 0.9479 - ETA: 0s - loss: 0.1752 - acc: 0.9476 - ETA: 0s - loss: 0.1753 - acc: 0.9471 - ETA: 0s - loss: 0.1748 - acc: 0.9471 - ETA: 0s - loss: 0.1745 - acc: 0.9473 - ETA: 0s - loss: 0.1731 - acc: 0.9477 - ETA: 0s - loss: 0.1727 - acc: 0.9479 - ETA: 0s - loss: 0.1719 - acc: 0.9481 - ETA: 0s - loss: 0.1723 - acc: 0.9478 - ETA: 0s - loss: 0.1723 - acc: 0.9475 - ETA: 0s - loss: 0.1717 - acc: 0.9475 - ETA: 0s - loss: 0.1708 - acc: 0.9477 - ETA: 0s - loss: 0.1740 - acc: 0.9469 - ETA: 0s - loss: 0.1750 - acc: 0.9469 - ETA: 0s - loss: 0.1764 - acc: 0.9464 - ETA: 0s - loss: 0.1774 - acc: 0.9458Epoch 00007: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1770 - acc: 0.9460 - val_loss: 0.5708 - val_acc: 0.8515
    Epoch 9/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.4235 - acc: 0.8500 - ETA: 6s - loss: 0.1940 - acc: 0.9375 - ETA: 6s - loss: 0.1980 - acc: 0.9071 - ETA: 6s - loss: 0.2069 - acc: 0.9100 - ETA: 6s - loss: 0.1829 - acc: 0.9231 - ETA: 6s - loss: 0.1802 - acc: 0.9281 - ETA: 6s - loss: 0.1602 - acc: 0.9395 - ETA: 6s - loss: 0.1432 - acc: 0.9477 - ETA: 6s - loss: 0.1530 - acc: 0.9460 - ETA: 6s - loss: 0.1527 - acc: 0.9464 - ETA: 6s - loss: 0.1464 - acc: 0.9484 - ETA: 6s - loss: 0.1479 - acc: 0.9471 - ETA: 6s - loss: 0.1546 - acc: 0.9486 - ETA: 6s - loss: 0.1480 - acc: 0.9513 - ETA: 6s - loss: 0.1531 - acc: 0.9476 - ETA: 6s - loss: 0.1602 - acc: 0.9467 - ETA: 6s - loss: 0.1599 - acc: 0.9479 - ETA: 6s - loss: 0.1702 - acc: 0.9461 - ETA: 6s - loss: 0.1693 - acc: 0.9472 - ETA: 5s - loss: 0.1698 - acc: 0.9474 - ETA: 5s - loss: 0.1683 - acc: 0.9467 - ETA: 5s - loss: 0.1621 - acc: 0.9492 - ETA: 5s - loss: 0.1583 - acc: 0.9508 - ETA: 5s - loss: 0.1668 - acc: 0.9493 - ETA: 5s - loss: 0.1616 - acc: 0.9507 - ETA: 5s - loss: 0.1578 - acc: 0.9513 - ETA: 5s - loss: 0.1542 - acc: 0.9519 - ETA: 5s - loss: 0.1573 - acc: 0.9506 - ETA: 5s - loss: 0.1578 - acc: 0.9506 - ETA: 5s - loss: 0.1562 - acc: 0.9517 - ETA: 5s - loss: 0.1651 - acc: 0.9506 - ETA: 5s - loss: 0.1609 - acc: 0.9516 - ETA: 5s - loss: 0.1566 - acc: 0.9531 - ETA: 5s - loss: 0.1565 - acc: 0.9530 - ETA: 4s - loss: 0.1555 - acc: 0.9529 - ETA: 4s - loss: 0.1544 - acc: 0.9533 - ETA: 4s - loss: 0.1568 - acc: 0.9532 - ETA: 4s - loss: 0.1540 - acc: 0.9536 - ETA: 4s - loss: 0.1510 - acc: 0.9544 - ETA: 4s - loss: 0.1505 - acc: 0.9547 - ETA: 4s - loss: 0.1500 - acc: 0.9550 - ETA: 4s - loss: 0.1476 - acc: 0.9557 - ETA: 4s - loss: 0.1468 - acc: 0.9560 - ETA: 4s - loss: 0.1469 - acc: 0.9562 - ETA: 4s - loss: 0.1502 - acc: 0.9553 - ETA: 4s - loss: 0.1488 - acc: 0.9560 - ETA: 4s - loss: 0.1506 - acc: 0.9551 - ETA: 4s - loss: 0.1521 - acc: 0.9546 - ETA: 4s - loss: 0.1528 - acc: 0.9545 - ETA: 3s - loss: 0.1521 - acc: 0.9538 - ETA: 3s - loss: 0.1537 - acc: 0.9530 - ETA: 3s - loss: 0.1541 - acc: 0.9530 - ETA: 3s - loss: 0.1551 - acc: 0.9532 - ETA: 3s - loss: 0.1542 - acc: 0.9535 - ETA: 3s - loss: 0.1528 - acc: 0.9540 - ETA: 3s - loss: 0.1559 - acc: 0.9534 - ETA: 3s - loss: 0.1548 - acc: 0.9533 - ETA: 3s - loss: 0.1563 - acc: 0.9529 - ETA: 3s - loss: 0.1570 - acc: 0.9529 - ETA: 3s - loss: 0.1610 - acc: 0.9520 - ETA: 3s - loss: 0.1602 - acc: 0.9522 - ETA: 3s - loss: 0.1616 - acc: 0.9514 - ETA: 3s - loss: 0.1611 - acc: 0.9514 - ETA: 3s - loss: 0.1630 - acc: 0.9508 - ETA: 3s - loss: 0.1644 - acc: 0.9508 - ETA: 2s - loss: 0.1639 - acc: 0.9510 - ETA: 2s - loss: 0.1621 - acc: 0.9515 - ETA: 2s - loss: 0.1650 - acc: 0.9510 - ETA: 2s - loss: 0.1691 - acc: 0.9507 - ETA: 2s - loss: 0.1675 - acc: 0.9510 - ETA: 2s - loss: 0.1668 - acc: 0.9507 - ETA: 2s - loss: 0.1690 - acc: 0.9498 - ETA: 2s - loss: 0.1672 - acc: 0.9502 - ETA: 2s - loss: 0.1676 - acc: 0.9498 - ETA: 2s - loss: 0.1702 - acc: 0.9498 - ETA: 2s - loss: 0.1697 - acc: 0.9500 - ETA: 2s - loss: 0.1684 - acc: 0.9504 - ETA: 2s - loss: 0.1702 - acc: 0.9502 - ETA: 2s - loss: 0.1732 - acc: 0.9496 - ETA: 2s - loss: 0.1715 - acc: 0.9498 - ETA: 2s - loss: 0.1714 - acc: 0.9498 - ETA: 1s - loss: 0.1696 - acc: 0.9504 - ETA: 1s - loss: 0.1682 - acc: 0.9508 - ETA: 1s - loss: 0.1668 - acc: 0.9510 - ETA: 1s - loss: 0.1665 - acc: 0.9510 - ETA: 1s - loss: 0.1654 - acc: 0.9512 - ETA: 1s - loss: 0.1644 - acc: 0.9514 - ETA: 1s - loss: 0.1666 - acc: 0.9506 - ETA: 1s - loss: 0.1671 - acc: 0.9502 - ETA: 1s - loss: 0.1666 - acc: 0.9502 - ETA: 1s - loss: 0.1662 - acc: 0.9502 - ETA: 1s - loss: 0.1660 - acc: 0.9504 - ETA: 1s - loss: 0.1652 - acc: 0.9505 - ETA: 1s - loss: 0.1651 - acc: 0.9507 - ETA: 1s - loss: 0.1650 - acc: 0.9507 - ETA: 1s - loss: 0.1637 - acc: 0.9512 - ETA: 1s - loss: 0.1631 - acc: 0.9510 - ETA: 0s - loss: 0.1631 - acc: 0.9510 - ETA: 0s - loss: 0.1634 - acc: 0.9509 - ETA: 0s - loss: 0.1647 - acc: 0.9502 - ETA: 0s - loss: 0.1650 - acc: 0.9498 - ETA: 0s - loss: 0.1657 - acc: 0.9492 - ETA: 0s - loss: 0.1649 - acc: 0.9495 - ETA: 0s - loss: 0.1647 - acc: 0.9493 - ETA: 0s - loss: 0.1643 - acc: 0.9490 - ETA: 0s - loss: 0.1689 - acc: 0.9481 - ETA: 0s - loss: 0.1689 - acc: 0.9481 - ETA: 0s - loss: 0.1677 - acc: 0.9484 - ETA: 0s - loss: 0.1678 - acc: 0.9484 - ETA: 0s - loss: 0.1673 - acc: 0.9485 - ETA: 0s - loss: 0.1660 - acc: 0.9489 - ETA: 0s - loss: 0.1649 - acc: 0.9494Epoch 00008: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1667 - acc: 0.9488 - val_loss: 0.5592 - val_acc: 0.8707
    Epoch 10/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.1232 - acc: 0.9500 - ETA: 6s - loss: 0.1668 - acc: 0.9375 - ETA: 6s - loss: 0.1685 - acc: 0.9429 - ETA: 6s - loss: 0.1514 - acc: 0.9500 - ETA: 6s - loss: 0.2203 - acc: 0.9458 - ETA: 6s - loss: 0.2183 - acc: 0.9467 - ETA: 6s - loss: 0.1877 - acc: 0.9528 - ETA: 6s - loss: 0.1790 - acc: 0.9524 - ETA: 6s - loss: 0.1737 - acc: 0.9521 - ETA: 6s - loss: 0.1653 - acc: 0.9556 - ETA: 6s - loss: 0.1760 - acc: 0.9583 - ETA: 6s - loss: 0.1689 - acc: 0.9576 - ETA: 6s - loss: 0.1645 - acc: 0.9583 - ETA: 6s - loss: 0.1560 - acc: 0.9590 - ETA: 6s - loss: 0.1647 - acc: 0.9583 - ETA: 6s - loss: 0.1861 - acc: 0.9556 - ETA: 6s - loss: 0.1766 - acc: 0.9583 - ETA: 6s - loss: 0.1777 - acc: 0.9588 - ETA: 5s - loss: 0.1750 - acc: 0.9583 - ETA: 5s - loss: 0.1740 - acc: 0.9579 - ETA: 5s - loss: 0.1791 - acc: 0.9567 - ETA: 5s - loss: 0.1733 - acc: 0.9579 - ETA: 5s - loss: 0.1671 - acc: 0.9598 - ETA: 5s - loss: 0.1678 - acc: 0.9594 - ETA: 5s - loss: 0.1729 - acc: 0.9569 - ETA: 5s - loss: 0.1715 - acc: 0.9580 - ETA: 5s - loss: 0.1741 - acc: 0.9564 - ETA: 5s - loss: 0.1715 - acc: 0.9562 - ETA: 5s - loss: 0.1674 - acc: 0.9571 - ETA: 5s - loss: 0.1678 - acc: 0.9569 - ETA: 5s - loss: 0.1682 - acc: 0.9567 - ETA: 5s - loss: 0.1666 - acc: 0.9565 - ETA: 5s - loss: 0.1657 - acc: 0.9568 - ETA: 5s - loss: 0.1664 - acc: 0.9556 - ETA: 4s - loss: 0.1724 - acc: 0.9534 - ETA: 4s - loss: 0.1699 - acc: 0.9543 - ETA: 4s - loss: 0.1684 - acc: 0.9542 - ETA: 4s - loss: 0.1665 - acc: 0.9545 - ETA: 4s - loss: 0.1701 - acc: 0.9535 - ETA: 4s - loss: 0.1667 - acc: 0.9547 - ETA: 4s - loss: 0.1660 - acc: 0.9546 - ETA: 4s - loss: 0.1675 - acc: 0.9545 - ETA: 4s - loss: 0.1647 - acc: 0.9556 - ETA: 4s - loss: 0.1634 - acc: 0.9554 - ETA: 4s - loss: 0.1610 - acc: 0.9561 - ETA: 4s - loss: 0.1598 - acc: 0.9567 - ETA: 4s - loss: 0.1596 - acc: 0.9569 - ETA: 4s - loss: 0.1597 - acc: 0.9567 - ETA: 4s - loss: 0.1598 - acc: 0.9566 - ETA: 4s - loss: 0.1600 - acc: 0.9561 - ETA: 3s - loss: 0.1594 - acc: 0.9564 - ETA: 3s - loss: 0.1600 - acc: 0.9556 - ETA: 3s - loss: 0.1583 - acc: 0.9561 - ETA: 3s - loss: 0.1601 - acc: 0.9554 - ETA: 3s - loss: 0.1584 - acc: 0.9556 - ETA: 3s - loss: 0.1596 - acc: 0.9552 - ETA: 3s - loss: 0.1593 - acc: 0.9551 - ETA: 3s - loss: 0.1596 - acc: 0.9547 - ETA: 3s - loss: 0.1573 - acc: 0.9552 - ETA: 3s - loss: 0.1567 - acc: 0.9551 - ETA: 3s - loss: 0.1565 - acc: 0.9550 - ETA: 3s - loss: 0.1574 - acc: 0.9547 - ETA: 3s - loss: 0.1555 - acc: 0.9551 - ETA: 3s - loss: 0.1541 - acc: 0.9556 - ETA: 3s - loss: 0.1535 - acc: 0.9550 - ETA: 2s - loss: 0.1526 - acc: 0.9552 - ETA: 2s - loss: 0.1523 - acc: 0.9554 - ETA: 2s - loss: 0.1530 - acc: 0.9555 - ETA: 2s - loss: 0.1540 - acc: 0.9554 - ETA: 2s - loss: 0.1548 - acc: 0.9554 - ETA: 2s - loss: 0.1545 - acc: 0.9555 - ETA: 2s - loss: 0.1559 - acc: 0.9545 - ETA: 2s - loss: 0.1551 - acc: 0.9549 - ETA: 2s - loss: 0.1564 - acc: 0.9541 - ETA: 2s - loss: 0.1572 - acc: 0.9541 - ETA: 2s - loss: 0.1568 - acc: 0.9543 - ETA: 2s - loss: 0.1580 - acc: 0.9542 - ETA: 2s - loss: 0.1584 - acc: 0.9539 - ETA: 2s - loss: 0.1579 - acc: 0.9539 - ETA: 2s - loss: 0.1569 - acc: 0.9540 - ETA: 2s - loss: 0.1583 - acc: 0.9538 - ETA: 1s - loss: 0.1578 - acc: 0.9537 - ETA: 1s - loss: 0.1578 - acc: 0.9533 - ETA: 1s - loss: 0.1574 - acc: 0.9536 - ETA: 1s - loss: 0.1558 - acc: 0.9542 - ETA: 1s - loss: 0.1548 - acc: 0.9545 - ETA: 1s - loss: 0.1546 - acc: 0.9545 - ETA: 1s - loss: 0.1537 - acc: 0.9548 - ETA: 1s - loss: 0.1526 - acc: 0.9552 - ETA: 1s - loss: 0.1514 - acc: 0.9557 - ETA: 1s - loss: 0.1502 - acc: 0.9558 - ETA: 1s - loss: 0.1499 - acc: 0.9554 - ETA: 1s - loss: 0.1531 - acc: 0.9547 - ETA: 1s - loss: 0.1522 - acc: 0.9551 - ETA: 1s - loss: 0.1523 - acc: 0.9548 - ETA: 1s - loss: 0.1525 - acc: 0.9546 - ETA: 1s - loss: 0.1515 - acc: 0.9549 - ETA: 0s - loss: 0.1519 - acc: 0.9543 - ETA: 0s - loss: 0.1517 - acc: 0.9543 - ETA: 0s - loss: 0.1507 - acc: 0.9547 - ETA: 0s - loss: 0.1507 - acc: 0.9549 - ETA: 0s - loss: 0.1498 - acc: 0.9550 - ETA: 0s - loss: 0.1485 - acc: 0.9554 - ETA: 0s - loss: 0.1484 - acc: 0.9555 - ETA: 0s - loss: 0.1475 - acc: 0.9556 - ETA: 0s - loss: 0.1473 - acc: 0.9559 - ETA: 0s - loss: 0.1477 - acc: 0.9559 - ETA: 0s - loss: 0.1472 - acc: 0.9558 - ETA: 0s - loss: 0.1498 - acc: 0.9557 - ETA: 0s - loss: 0.1494 - acc: 0.9557 - ETA: 0s - loss: 0.1494 - acc: 0.9556 - ETA: 0s - loss: 0.1484 - acc: 0.9559Epoch 00009: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1501 - acc: 0.9558 - val_loss: 0.5725 - val_acc: 0.8479
    Epoch 11/20
    6660/6680 [============================>.] - ETA: 7s - loss: 0.0180 - acc: 1.0000 - ETA: 6s - loss: 0.0517 - acc: 0.9750 - ETA: 6s - loss: 0.0865 - acc: 0.9714 - ETA: 6s - loss: 0.0666 - acc: 0.9750 - ETA: 6s - loss: 0.0605 - acc: 0.9769 - ETA: 6s - loss: 0.1065 - acc: 0.9687 - ETA: 6s - loss: 0.1143 - acc: 0.9684 - ETA: 6s - loss: 0.1107 - acc: 0.9682 - ETA: 6s - loss: 0.1245 - acc: 0.9640 - ETA: 6s - loss: 0.1167 - acc: 0.9643 - ETA: 6s - loss: 0.1147 - acc: 0.9661 - ETA: 6s - loss: 0.1059 - acc: 0.9691 - ETA: 6s - loss: 0.1012 - acc: 0.9703 - ETA: 6s - loss: 0.0949 - acc: 0.9725 - ETA: 6s - loss: 0.0951 - acc: 0.9709 - ETA: 6s - loss: 0.0910 - acc: 0.9717 - ETA: 6s - loss: 0.0940 - acc: 0.9714 - ETA: 6s - loss: 0.0906 - acc: 0.9731 - ETA: 5s - loss: 0.0936 - acc: 0.9736 - ETA: 5s - loss: 0.0990 - acc: 0.9724 - ETA: 5s - loss: 0.0975 - acc: 0.9721 - ETA: 5s - loss: 0.0999 - acc: 0.9703 - ETA: 5s - loss: 0.0968 - acc: 0.9716 - ETA: 5s - loss: 0.0943 - acc: 0.9721 - ETA: 5s - loss: 0.0938 - acc: 0.9719 - ETA: 5s - loss: 0.0947 - acc: 0.9717 - ETA: 5s - loss: 0.0952 - acc: 0.9715 - ETA: 5s - loss: 0.0940 - acc: 0.9720 - ETA: 5s - loss: 0.0938 - acc: 0.9718 - ETA: 5s - loss: 0.0976 - acc: 0.9705 - ETA: 5s - loss: 0.0960 - acc: 0.9709 - ETA: 5s - loss: 0.0951 - acc: 0.9702 - ETA: 4s - loss: 0.0958 - acc: 0.9706 - ETA: 4s - loss: 0.1045 - acc: 0.9690 - ETA: 4s - loss: 0.1071 - acc: 0.9680 - ETA: 4s - loss: 0.1052 - acc: 0.9684 - ETA: 4s - loss: 0.1049 - acc: 0.9683 - ETA: 4s - loss: 0.1060 - acc: 0.9683 - ETA: 4s - loss: 0.1076 - acc: 0.9683 - ETA: 4s - loss: 0.1103 - acc: 0.9674 - ETA: 4s - loss: 0.1124 - acc: 0.9661 - ETA: 4s - loss: 0.1129 - acc: 0.9657 - ETA: 4s - loss: 0.1114 - acc: 0.9661 - ETA: 4s - loss: 0.1119 - acc: 0.9662 - ETA: 4s - loss: 0.1109 - acc: 0.9662 - ETA: 4s - loss: 0.1095 - acc: 0.9665 - ETA: 4s - loss: 0.1091 - acc: 0.9665 - ETA: 4s - loss: 0.1089 - acc: 0.9662 - ETA: 3s - loss: 0.1150 - acc: 0.9655 - ETA: 3s - loss: 0.1153 - acc: 0.9652 - ETA: 3s - loss: 0.1201 - acc: 0.9642 - ETA: 3s - loss: 0.1236 - acc: 0.9640 - ETA: 3s - loss: 0.1248 - acc: 0.9634 - ETA: 3s - loss: 0.1254 - acc: 0.9634 - ETA: 3s - loss: 0.1256 - acc: 0.9632 - ETA: 3s - loss: 0.1245 - acc: 0.9636 - ETA: 3s - loss: 0.1258 - acc: 0.9627 - ETA: 3s - loss: 0.1261 - acc: 0.9625 - ETA: 3s - loss: 0.1256 - acc: 0.9629 - ETA: 3s - loss: 0.1260 - acc: 0.9626 - ETA: 3s - loss: 0.1271 - acc: 0.9619 - ETA: 3s - loss: 0.1266 - acc: 0.9620 - ETA: 3s - loss: 0.1287 - acc: 0.9615 - ETA: 3s - loss: 0.1298 - acc: 0.9613 - ETA: 2s - loss: 0.1315 - acc: 0.9606 - ETA: 2s - loss: 0.1381 - acc: 0.9597 - ETA: 2s - loss: 0.1402 - acc: 0.9590 - ETA: 2s - loss: 0.1387 - acc: 0.9594 - ETA: 2s - loss: 0.1383 - acc: 0.9595 - ETA: 2s - loss: 0.1400 - acc: 0.9596 - ETA: 2s - loss: 0.1435 - acc: 0.9595 - ETA: 2s - loss: 0.1425 - acc: 0.9596 - ETA: 2s - loss: 0.1411 - acc: 0.9599 - ETA: 2s - loss: 0.1411 - acc: 0.9598 - ETA: 2s - loss: 0.1403 - acc: 0.9599 - ETA: 2s - loss: 0.1404 - acc: 0.9595 - ETA: 2s - loss: 0.1401 - acc: 0.9594 - ETA: 2s - loss: 0.1389 - acc: 0.9597 - ETA: 2s - loss: 0.1389 - acc: 0.9598 - ETA: 2s - loss: 0.1388 - acc: 0.9597 - ETA: 1s - loss: 0.1384 - acc: 0.9598 - ETA: 1s - loss: 0.1412 - acc: 0.9586 - ETA: 1s - loss: 0.1399 - acc: 0.9589 - ETA: 1s - loss: 0.1417 - acc: 0.9588 - ETA: 1s - loss: 0.1412 - acc: 0.9591 - ETA: 1s - loss: 0.1413 - acc: 0.9590 - ETA: 1s - loss: 0.1411 - acc: 0.9591 - ETA: 1s - loss: 0.1412 - acc: 0.9590 - ETA: 1s - loss: 0.1404 - acc: 0.9593 - ETA: 1s - loss: 0.1406 - acc: 0.9590 - ETA: 1s - loss: 0.1397 - acc: 0.9593 - ETA: 1s - loss: 0.1393 - acc: 0.9590 - ETA: 1s - loss: 0.1396 - acc: 0.9591 - ETA: 1s - loss: 0.1388 - acc: 0.9591 - ETA: 1s - loss: 0.1385 - acc: 0.9592 - ETA: 1s - loss: 0.1384 - acc: 0.9591 - ETA: 0s - loss: 0.1384 - acc: 0.9594 - ETA: 0s - loss: 0.1379 - acc: 0.9595 - ETA: 0s - loss: 0.1378 - acc: 0.9594 - ETA: 0s - loss: 0.1392 - acc: 0.9589 - ETA: 0s - loss: 0.1410 - acc: 0.9588 - ETA: 0s - loss: 0.1409 - acc: 0.9589 - ETA: 0s - loss: 0.1407 - acc: 0.9588 - ETA: 0s - loss: 0.1407 - acc: 0.9587 - ETA: 0s - loss: 0.1400 - acc: 0.9590 - ETA: 0s - loss: 0.1395 - acc: 0.9590 - ETA: 0s - loss: 0.1387 - acc: 0.9593 - ETA: 0s - loss: 0.1384 - acc: 0.9593 - ETA: 0s - loss: 0.1376 - acc: 0.9596 - ETA: 0s - loss: 0.1386 - acc: 0.9593 - ETA: 0s - loss: 0.1393 - acc: 0.9595 - ETA: 0s - loss: 0.1393 - acc: 0.9593Epoch 00010: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1389 - acc: 0.9594 - val_loss: 0.5652 - val_acc: 0.8491
    Epoch 12/20
    6640/6680 [============================>.] - ETA: 6s - loss: 0.0324 - acc: 1.0000 - ETA: 6s - loss: 0.1854 - acc: 0.9750 - ETA: 6s - loss: 0.1316 - acc: 0.9714 - ETA: 6s - loss: 0.1121 - acc: 0.9700 - ETA: 6s - loss: 0.0965 - acc: 0.9731 - ETA: 6s - loss: 0.0825 - acc: 0.9781 - ETA: 6s - loss: 0.0777 - acc: 0.9789 - ETA: 6s - loss: 0.1003 - acc: 0.9773 - ETA: 6s - loss: 0.0970 - acc: 0.9740 - ETA: 6s - loss: 0.0952 - acc: 0.9750 - ETA: 6s - loss: 0.0967 - acc: 0.9742 - ETA: 6s - loss: 0.1112 - acc: 0.9662 - ETA: 5s - loss: 0.1134 - acc: 0.9635 - ETA: 5s - loss: 0.1100 - acc: 0.9650 - ETA: 5s - loss: 0.1108 - acc: 0.9663 - ETA: 5s - loss: 0.1111 - acc: 0.9663 - ETA: 5s - loss: 0.1155 - acc: 0.9653 - ETA: 5s - loss: 0.1118 - acc: 0.9663 - ETA: 5s - loss: 0.1111 - acc: 0.9664 - ETA: 5s - loss: 0.1060 - acc: 0.9681 - ETA: 5s - loss: 0.1145 - acc: 0.9656 - ETA: 5s - loss: 0.1131 - acc: 0.9656 - ETA: 5s - loss: 0.1120 - acc: 0.9642 - ETA: 5s - loss: 0.1116 - acc: 0.9636 - ETA: 5s - loss: 0.1177 - acc: 0.9630 - ETA: 5s - loss: 0.1169 - acc: 0.9632 - ETA: 5s - loss: 0.1140 - acc: 0.9639 - ETA: 5s - loss: 0.1120 - acc: 0.9646 - ETA: 5s - loss: 0.1163 - acc: 0.9635 - ETA: 4s - loss: 0.1222 - acc: 0.9631 - ETA: 4s - loss: 0.1217 - acc: 0.9626 - ETA: 4s - loss: 0.1226 - acc: 0.9622 - ETA: 4s - loss: 0.1197 - acc: 0.9634 - ETA: 4s - loss: 0.1170 - acc: 0.9640 - ETA: 4s - loss: 0.1182 - acc: 0.9636 - ETA: 4s - loss: 0.1163 - acc: 0.9642 - ETA: 4s - loss: 0.1142 - acc: 0.9647 - ETA: 4s - loss: 0.1192 - acc: 0.9643 - ETA: 4s - loss: 0.1183 - acc: 0.9648 - ETA: 4s - loss: 0.1194 - acc: 0.9644 - ETA: 4s - loss: 0.1183 - acc: 0.9649 - ETA: 4s - loss: 0.1194 - acc: 0.9645 - ETA: 4s - loss: 0.1191 - acc: 0.9646 - ETA: 4s - loss: 0.1196 - acc: 0.9642 - ETA: 4s - loss: 0.1208 - acc: 0.9632 - ETA: 4s - loss: 0.1209 - acc: 0.9632 - ETA: 4s - loss: 0.1236 - acc: 0.9622 - ETA: 3s - loss: 0.1249 - acc: 0.9616 - ETA: 3s - loss: 0.1264 - acc: 0.9610 - ETA: 3s - loss: 0.1254 - acc: 0.9615 - ETA: 3s - loss: 0.1241 - acc: 0.9616 - ETA: 3s - loss: 0.1265 - acc: 0.9604 - ETA: 3s - loss: 0.1254 - acc: 0.9605 - ETA: 3s - loss: 0.1243 - acc: 0.9609 - ETA: 3s - loss: 0.1242 - acc: 0.9607 - ETA: 3s - loss: 0.1248 - acc: 0.9602 - ETA: 3s - loss: 0.1270 - acc: 0.9598 - ETA: 3s - loss: 0.1308 - acc: 0.9587 - ETA: 3s - loss: 0.1300 - acc: 0.9586 - ETA: 3s - loss: 0.1300 - acc: 0.9584 - ETA: 3s - loss: 0.1299 - acc: 0.9583 - ETA: 3s - loss: 0.1286 - acc: 0.9587 - ETA: 3s - loss: 0.1272 - acc: 0.9591 - ETA: 3s - loss: 0.1261 - acc: 0.9595 - ETA: 2s - loss: 0.1263 - acc: 0.9591 - ETA: 2s - loss: 0.1274 - acc: 0.9590 - ETA: 2s - loss: 0.1260 - acc: 0.9593 - ETA: 2s - loss: 0.1255 - acc: 0.9595 - ETA: 2s - loss: 0.1239 - acc: 0.9600 - ETA: 2s - loss: 0.1284 - acc: 0.9587 - ETA: 2s - loss: 0.1324 - acc: 0.9583 - ETA: 2s - loss: 0.1327 - acc: 0.9585 - ETA: 2s - loss: 0.1328 - acc: 0.9586 - ETA: 2s - loss: 0.1327 - acc: 0.9587 - ETA: 2s - loss: 0.1367 - acc: 0.9577 - ETA: 2s - loss: 0.1359 - acc: 0.9580 - ETA: 2s - loss: 0.1352 - acc: 0.9581 - ETA: 2s - loss: 0.1348 - acc: 0.9584 - ETA: 2s - loss: 0.1359 - acc: 0.9583 - ETA: 2s - loss: 0.1347 - acc: 0.9586 - ETA: 1s - loss: 0.1334 - acc: 0.9592 - ETA: 1s - loss: 0.1339 - acc: 0.9593 - ETA: 1s - loss: 0.1333 - acc: 0.9596 - ETA: 1s - loss: 0.1343 - acc: 0.9590 - ETA: 1s - loss: 0.1340 - acc: 0.9593 - ETA: 1s - loss: 0.1338 - acc: 0.9592 - ETA: 1s - loss: 0.1325 - acc: 0.9597 - ETA: 1s - loss: 0.1312 - acc: 0.9602 - ETA: 1s - loss: 0.1301 - acc: 0.9604 - ETA: 1s - loss: 0.1299 - acc: 0.9605 - ETA: 1s - loss: 0.1289 - acc: 0.9607 - ETA: 1s - loss: 0.1297 - acc: 0.9604 - ETA: 1s - loss: 0.1305 - acc: 0.9603 - ETA: 1s - loss: 0.1311 - acc: 0.9602 - ETA: 1s - loss: 0.1332 - acc: 0.9598 - ETA: 1s - loss: 0.1332 - acc: 0.9597 - ETA: 0s - loss: 0.1324 - acc: 0.9598 - ETA: 0s - loss: 0.1313 - acc: 0.9602 - ETA: 0s - loss: 0.1305 - acc: 0.9602 - ETA: 0s - loss: 0.1303 - acc: 0.9605 - ETA: 0s - loss: 0.1297 - acc: 0.9605 - ETA: 0s - loss: 0.1288 - acc: 0.9606 - ETA: 0s - loss: 0.1283 - acc: 0.9607 - ETA: 0s - loss: 0.1290 - acc: 0.9606 - ETA: 0s - loss: 0.1284 - acc: 0.9606 - ETA: 0s - loss: 0.1290 - acc: 0.9604 - ETA: 0s - loss: 0.1288 - acc: 0.9606 - ETA: 0s - loss: 0.1295 - acc: 0.9605 - ETA: 0s - loss: 0.1305 - acc: 0.9602 - ETA: 0s - loss: 0.1304 - acc: 0.9603 - ETA: 0s - loss: 0.1299 - acc: 0.9605 - ETA: 0s - loss: 0.1293 - acc: 0.9607Epoch 00011: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1306 - acc: 0.9605 - val_loss: 0.5974 - val_acc: 0.8587
    Epoch 13/20
    6660/6680 [============================>.] - ETA: 6s - loss: 0.0423 - acc: 1.0000 - ETA: 6s - loss: 0.0686 - acc: 0.9875 - ETA: 6s - loss: 0.0861 - acc: 0.9786 - ETA: 6s - loss: 0.0701 - acc: 0.9800 - ETA: 6s - loss: 0.0660 - acc: 0.9769 - ETA: 6s - loss: 0.0736 - acc: 0.9781 - ETA: 6s - loss: 0.0655 - acc: 0.9816 - ETA: 6s - loss: 0.0687 - acc: 0.9818 - ETA: 6s - loss: 0.0663 - acc: 0.9800 - ETA: 6s - loss: 0.0656 - acc: 0.9786 - ETA: 6s - loss: 0.0694 - acc: 0.9774 - ETA: 6s - loss: 0.0656 - acc: 0.9794 - ETA: 6s - loss: 0.0640 - acc: 0.9797 - ETA: 6s - loss: 0.0824 - acc: 0.9762 - ETA: 6s - loss: 0.0809 - acc: 0.9767 - ETA: 5s - loss: 0.0919 - acc: 0.9717 - ETA: 5s - loss: 0.0938 - acc: 0.9704 - ETA: 5s - loss: 0.0968 - acc: 0.9702 - ETA: 5s - loss: 0.0991 - acc: 0.9700 - ETA: 5s - loss: 0.0970 - acc: 0.9707 - ETA: 5s - loss: 0.0969 - acc: 0.9705 - ETA: 5s - loss: 0.0937 - acc: 0.9711 - ETA: 5s - loss: 0.0921 - acc: 0.9716 - ETA: 5s - loss: 0.0919 - acc: 0.9721 - ETA: 5s - loss: 0.0929 - acc: 0.9712 - ETA: 5s - loss: 0.0980 - acc: 0.9704 - ETA: 5s - loss: 0.0961 - acc: 0.9709 - ETA: 5s - loss: 0.0950 - acc: 0.9713 - ETA: 5s - loss: 0.0961 - acc: 0.9706 - ETA: 5s - loss: 0.0995 - acc: 0.9693 - ETA: 5s - loss: 0.1005 - acc: 0.9681 - ETA: 5s - loss: 0.1032 - acc: 0.9676 - ETA: 4s - loss: 0.1057 - acc: 0.9660 - ETA: 4s - loss: 0.1070 - acc: 0.9655 - ETA: 4s - loss: 0.1073 - acc: 0.9650 - ETA: 4s - loss: 0.1047 - acc: 0.9660 - ETA: 4s - loss: 0.1072 - acc: 0.9656 - ETA: 4s - loss: 0.1061 - acc: 0.9661 - ETA: 4s - loss: 0.1059 - acc: 0.9661 - ETA: 4s - loss: 0.1105 - acc: 0.9657 - ETA: 4s - loss: 0.1133 - acc: 0.9649 - ETA: 4s - loss: 0.1120 - acc: 0.9653 - ETA: 4s - loss: 0.1151 - acc: 0.9646 - ETA: 4s - loss: 0.1154 - acc: 0.9650 - ETA: 4s - loss: 0.1144 - acc: 0.9650 - ETA: 4s - loss: 0.1130 - acc: 0.9651 - ETA: 4s - loss: 0.1131 - acc: 0.9644 - ETA: 4s - loss: 0.1120 - acc: 0.9648 - ETA: 4s - loss: 0.1111 - acc: 0.9648 - ETA: 3s - loss: 0.1096 - acc: 0.9652 - ETA: 3s - loss: 0.1098 - acc: 0.9646 - ETA: 3s - loss: 0.1101 - acc: 0.9643 - ETA: 3s - loss: 0.1097 - acc: 0.9646 - ETA: 3s - loss: 0.1094 - acc: 0.9648 - ETA: 3s - loss: 0.1083 - acc: 0.9648 - ETA: 3s - loss: 0.1079 - acc: 0.9642 - ETA: 3s - loss: 0.1079 - acc: 0.9643 - ETA: 3s - loss: 0.1062 - acc: 0.9649 - ETA: 3s - loss: 0.1089 - acc: 0.9641 - ETA: 3s - loss: 0.1084 - acc: 0.9641 - ETA: 3s - loss: 0.1100 - acc: 0.9644 - ETA: 3s - loss: 0.1095 - acc: 0.9648 - ETA: 3s - loss: 0.1079 - acc: 0.9653 - ETA: 3s - loss: 0.1085 - acc: 0.9648 - ETA: 3s - loss: 0.1083 - acc: 0.9651 - ETA: 2s - loss: 0.1068 - acc: 0.9656 - ETA: 2s - loss: 0.1060 - acc: 0.9659 - ETA: 2s - loss: 0.1064 - acc: 0.9659 - ETA: 2s - loss: 0.1053 - acc: 0.9664 - ETA: 2s - loss: 0.1060 - acc: 0.9664 - ETA: 2s - loss: 0.1069 - acc: 0.9657 - ETA: 2s - loss: 0.1068 - acc: 0.9657 - ETA: 2s - loss: 0.1067 - acc: 0.9657 - ETA: 2s - loss: 0.1066 - acc: 0.9658 - ETA: 2s - loss: 0.1059 - acc: 0.9660 - ETA: 2s - loss: 0.1065 - acc: 0.9658 - ETA: 2s - loss: 0.1061 - acc: 0.9660 - ETA: 2s - loss: 0.1072 - acc: 0.9654 - ETA: 2s - loss: 0.1113 - acc: 0.9652 - ETA: 2s - loss: 0.1107 - acc: 0.9652 - ETA: 2s - loss: 0.1099 - acc: 0.9654 - ETA: 1s - loss: 0.1104 - acc: 0.9654 - ETA: 1s - loss: 0.1115 - acc: 0.9652 - ETA: 1s - loss: 0.1112 - acc: 0.9655 - ETA: 1s - loss: 0.1104 - acc: 0.9657 - ETA: 1s - loss: 0.1103 - acc: 0.9655 - ETA: 1s - loss: 0.1110 - acc: 0.9651 - ETA: 1s - loss: 0.1102 - acc: 0.9653 - ETA: 1s - loss: 0.1098 - acc: 0.9653 - ETA: 1s - loss: 0.1090 - acc: 0.9655 - ETA: 1s - loss: 0.1107 - acc: 0.9652 - ETA: 1s - loss: 0.1122 - acc: 0.9648 - ETA: 1s - loss: 0.1121 - acc: 0.9649 - ETA: 1s - loss: 0.1112 - acc: 0.9652 - ETA: 1s - loss: 0.1106 - acc: 0.9651 - ETA: 1s - loss: 0.1098 - acc: 0.9654 - ETA: 0s - loss: 0.1114 - acc: 0.9651 - ETA: 0s - loss: 0.1109 - acc: 0.9651 - ETA: 0s - loss: 0.1123 - acc: 0.9650 - ETA: 0s - loss: 0.1123 - acc: 0.9650 - ETA: 0s - loss: 0.1117 - acc: 0.9652 - ETA: 0s - loss: 0.1112 - acc: 0.9652 - ETA: 0s - loss: 0.1119 - acc: 0.9649 - ETA: 0s - loss: 0.1113 - acc: 0.9650 - ETA: 0s - loss: 0.1139 - acc: 0.9643 - ETA: 0s - loss: 0.1152 - acc: 0.9641 - ETA: 0s - loss: 0.1150 - acc: 0.9638 - ETA: 0s - loss: 0.1154 - acc: 0.9637 - ETA: 0s - loss: 0.1154 - acc: 0.9639 - ETA: 0s - loss: 0.1154 - acc: 0.9636 - ETA: 0s - loss: 0.1146 - acc: 0.9639 - ETA: 0s - loss: 0.1170 - acc: 0.9637Epoch 00012: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1169 - acc: 0.9638 - val_loss: 0.6100 - val_acc: 0.8551
    Epoch 14/20
    6640/6680 [============================>.] - ETA: 6s - loss: 0.0154 - acc: 1.0000 - ETA: 6s - loss: 0.1874 - acc: 0.9500 - ETA: 6s - loss: 0.1601 - acc: 0.9571 - ETA: 6s - loss: 0.1212 - acc: 0.9650 - ETA: 6s - loss: 0.1260 - acc: 0.9692 - ETA: 6s - loss: 0.1382 - acc: 0.9719 - ETA: 6s - loss: 0.1313 - acc: 0.9711 - ETA: 6s - loss: 0.1292 - acc: 0.9705 - ETA: 6s - loss: 0.1158 - acc: 0.9740 - ETA: 6s - loss: 0.1193 - acc: 0.9696 - ETA: 6s - loss: 0.1136 - acc: 0.9694 - ETA: 6s - loss: 0.1074 - acc: 0.9691 - ETA: 6s - loss: 0.1118 - acc: 0.9676 - ETA: 5s - loss: 0.1095 - acc: 0.9662 - ETA: 5s - loss: 0.1160 - acc: 0.9640 - ETA: 5s - loss: 0.1140 - acc: 0.9641 - ETA: 5s - loss: 0.1141 - acc: 0.9653 - ETA: 5s - loss: 0.1085 - acc: 0.9673 - ETA: 5s - loss: 0.1056 - acc: 0.9682 - ETA: 5s - loss: 0.1129 - acc: 0.9664 - ETA: 5s - loss: 0.1120 - acc: 0.9656 - ETA: 5s - loss: 0.1119 - acc: 0.9648 - ETA: 5s - loss: 0.1091 - acc: 0.9649 - ETA: 5s - loss: 0.1081 - acc: 0.9657 - ETA: 5s - loss: 0.1076 - acc: 0.9658 - ETA: 5s - loss: 0.1061 - acc: 0.9664 - ETA: 5s - loss: 0.1043 - acc: 0.9665 - ETA: 5s - loss: 0.1026 - acc: 0.9665 - ETA: 5s - loss: 0.0996 - acc: 0.9676 - ETA: 4s - loss: 0.1005 - acc: 0.9676 - ETA: 4s - loss: 0.1020 - acc: 0.9670 - ETA: 4s - loss: 0.1035 - acc: 0.9665 - ETA: 4s - loss: 0.1011 - acc: 0.9670 - ETA: 4s - loss: 0.1022 - acc: 0.9665 - ETA: 4s - loss: 0.1010 - acc: 0.9670 - ETA: 4s - loss: 0.0989 - acc: 0.9679 - ETA: 4s - loss: 0.0972 - acc: 0.9683 - ETA: 4s - loss: 0.0958 - acc: 0.9687 - ETA: 4s - loss: 0.0974 - acc: 0.9687 - ETA: 4s - loss: 0.0953 - acc: 0.9695 - ETA: 4s - loss: 0.0940 - acc: 0.9698 - ETA: 4s - loss: 0.0922 - acc: 0.9706 - ETA: 4s - loss: 0.0920 - acc: 0.9709 - ETA: 4s - loss: 0.0937 - acc: 0.9704 - ETA: 4s - loss: 0.0953 - acc: 0.9703 - ETA: 4s - loss: 0.0985 - acc: 0.9695 - ETA: 4s - loss: 0.0979 - acc: 0.9698 - ETA: 3s - loss: 0.0984 - acc: 0.9701 - ETA: 3s - loss: 0.0986 - acc: 0.9700 - ETA: 3s - loss: 0.0982 - acc: 0.9699 - ETA: 3s - loss: 0.0977 - acc: 0.9702 - ETA: 3s - loss: 0.0981 - acc: 0.9705 - ETA: 3s - loss: 0.0976 - acc: 0.9701 - ETA: 3s - loss: 0.0966 - acc: 0.9703 - ETA: 3s - loss: 0.0991 - acc: 0.9696 - ETA: 3s - loss: 0.0992 - acc: 0.9693 - ETA: 3s - loss: 0.1013 - acc: 0.9683 - ETA: 3s - loss: 0.1005 - acc: 0.9683 - ETA: 3s - loss: 0.1010 - acc: 0.9680 - ETA: 3s - loss: 0.1044 - acc: 0.9674 - ETA: 3s - loss: 0.1050 - acc: 0.9674 - ETA: 3s - loss: 0.1054 - acc: 0.9674 - ETA: 3s - loss: 0.1072 - acc: 0.9674 - ETA: 3s - loss: 0.1085 - acc: 0.9666 - ETA: 2s - loss: 0.1089 - acc: 0.9658 - ETA: 2s - loss: 0.1123 - acc: 0.9651 - ETA: 2s - loss: 0.1148 - acc: 0.9651 - ETA: 2s - loss: 0.1162 - acc: 0.9653 - ETA: 2s - loss: 0.1147 - acc: 0.9659 - ETA: 2s - loss: 0.1134 - acc: 0.9663 - ETA: 2s - loss: 0.1139 - acc: 0.9664 - ETA: 2s - loss: 0.1131 - acc: 0.9666 - ETA: 2s - loss: 0.1117 - acc: 0.9671 - ETA: 2s - loss: 0.1119 - acc: 0.9670 - ETA: 2s - loss: 0.1109 - acc: 0.9673 - ETA: 2s - loss: 0.1111 - acc: 0.9673 - ETA: 2s - loss: 0.1105 - acc: 0.9675 - ETA: 2s - loss: 0.1095 - acc: 0.9677 - ETA: 2s - loss: 0.1090 - acc: 0.9679 - ETA: 2s - loss: 0.1086 - acc: 0.9681 - ETA: 1s - loss: 0.1093 - acc: 0.9681 - ETA: 1s - loss: 0.1091 - acc: 0.9681 - ETA: 1s - loss: 0.1084 - acc: 0.9681 - ETA: 1s - loss: 0.1085 - acc: 0.9681 - ETA: 1s - loss: 0.1084 - acc: 0.9679 - ETA: 1s - loss: 0.1076 - acc: 0.9680 - ETA: 1s - loss: 0.1069 - acc: 0.9682 - ETA: 1s - loss: 0.1070 - acc: 0.9684 - ETA: 1s - loss: 0.1059 - acc: 0.9687 - ETA: 1s - loss: 0.1065 - acc: 0.9680 - ETA: 1s - loss: 0.1077 - acc: 0.9676 - ETA: 1s - loss: 0.1095 - acc: 0.9674 - ETA: 1s - loss: 0.1090 - acc: 0.9674 - ETA: 1s - loss: 0.1086 - acc: 0.9674 - ETA: 1s - loss: 0.1086 - acc: 0.9672 - ETA: 1s - loss: 0.1087 - acc: 0.9672 - ETA: 0s - loss: 0.1081 - acc: 0.9672 - ETA: 0s - loss: 0.1100 - acc: 0.9672 - ETA: 0s - loss: 0.1104 - acc: 0.9668 - ETA: 0s - loss: 0.1097 - acc: 0.9670 - ETA: 0s - loss: 0.1093 - acc: 0.9672 - ETA: 0s - loss: 0.1096 - acc: 0.9672 - ETA: 0s - loss: 0.1105 - acc: 0.9670 - ETA: 0s - loss: 0.1096 - acc: 0.9673 - ETA: 0s - loss: 0.1089 - acc: 0.9673 - ETA: 0s - loss: 0.1085 - acc: 0.9671 - ETA: 0s - loss: 0.1086 - acc: 0.9670 - ETA: 0s - loss: 0.1077 - acc: 0.9673 - ETA: 0s - loss: 0.1069 - acc: 0.9676 - ETA: 0s - loss: 0.1084 - acc: 0.9670 - ETA: 0s - loss: 0.1078 - acc: 0.9670 - ETA: 0s - loss: 0.1082 - acc: 0.9669Epoch 00013: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.1091 - acc: 0.9668 - val_loss: 0.6127 - val_acc: 0.8599
    Epoch 15/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.0152 - acc: 1.0000 - ETA: 6s - loss: 0.0877 - acc: 0.9750 - ETA: 6s - loss: 0.0689 - acc: 0.9786 - ETA: 6s - loss: 0.0548 - acc: 0.9850 - ETA: 6s - loss: 0.0625 - acc: 0.9846 - ETA: 6s - loss: 0.0563 - acc: 0.9844 - ETA: 6s - loss: 0.0512 - acc: 0.9868 - ETA: 6s - loss: 0.0488 - acc: 0.9864 - ETA: 6s - loss: 0.0442 - acc: 0.9880 - ETA: 6s - loss: 0.0495 - acc: 0.9857 - ETA: 6s - loss: 0.0514 - acc: 0.9855 - ETA: 6s - loss: 0.0571 - acc: 0.9824 - ETA: 6s - loss: 0.0607 - acc: 0.9824 - ETA: 5s - loss: 0.0582 - acc: 0.9837 - ETA: 5s - loss: 0.0716 - acc: 0.9814 - ETA: 5s - loss: 0.0728 - acc: 0.9804 - ETA: 5s - loss: 0.0749 - acc: 0.9806 - ETA: 5s - loss: 0.0741 - acc: 0.9808 - ETA: 5s - loss: 0.0707 - acc: 0.9818 - ETA: 5s - loss: 0.0696 - acc: 0.9819 - ETA: 5s - loss: 0.0720 - acc: 0.9820 - ETA: 5s - loss: 0.0739 - acc: 0.9812 - ETA: 5s - loss: 0.0728 - acc: 0.9806 - ETA: 5s - loss: 0.0718 - acc: 0.9807 - ETA: 5s - loss: 0.0718 - acc: 0.9808 - ETA: 5s - loss: 0.0726 - acc: 0.9803 - ETA: 5s - loss: 0.0744 - acc: 0.9791 - ETA: 5s - loss: 0.0856 - acc: 0.9762 - ETA: 5s - loss: 0.0835 - acc: 0.9771 - ETA: 4s - loss: 0.0827 - acc: 0.9773 - ETA: 4s - loss: 0.0814 - acc: 0.9775 - ETA: 4s - loss: 0.0802 - acc: 0.9777 - ETA: 4s - loss: 0.0828 - acc: 0.9768 - ETA: 4s - loss: 0.0859 - acc: 0.9760 - ETA: 4s - loss: 0.0867 - acc: 0.9752 - ETA: 4s - loss: 0.0865 - acc: 0.9755 - ETA: 4s - loss: 0.0864 - acc: 0.9752 - ETA: 4s - loss: 0.0868 - acc: 0.9746 - ETA: 4s - loss: 0.0875 - acc: 0.9743 - ETA: 4s - loss: 0.0871 - acc: 0.9742 - ETA: 4s - loss: 0.0877 - acc: 0.9744 - ETA: 4s - loss: 0.0875 - acc: 0.9742 - ETA: 4s - loss: 0.0888 - acc: 0.9740 - ETA: 4s - loss: 0.0916 - acc: 0.9738 - ETA: 4s - loss: 0.0900 - acc: 0.9744 - ETA: 3s - loss: 0.0927 - acc: 0.9739 - ETA: 3s - loss: 0.0922 - acc: 0.9737 - ETA: 3s - loss: 0.0939 - acc: 0.9729 - ETA: 3s - loss: 0.0924 - acc: 0.9734 - ETA: 3s - loss: 0.0909 - acc: 0.9740 - ETA: 3s - loss: 0.0903 - acc: 0.9742 - ETA: 3s - loss: 0.0895 - acc: 0.9740 - ETA: 3s - loss: 0.0882 - acc: 0.9745 - ETA: 3s - loss: 0.0891 - acc: 0.9744 - ETA: 3s - loss: 0.0898 - acc: 0.9739 - ETA: 3s - loss: 0.0904 - acc: 0.9738 - ETA: 3s - loss: 0.0893 - acc: 0.9743 - ETA: 3s - loss: 0.0923 - acc: 0.9738 - ETA: 3s - loss: 0.0919 - acc: 0.9737 - ETA: 3s - loss: 0.0910 - acc: 0.9739 - ETA: 3s - loss: 0.0948 - acc: 0.9729 - ETA: 3s - loss: 0.0951 - acc: 0.9726 - ETA: 2s - loss: 0.0950 - acc: 0.9725 - ETA: 2s - loss: 0.0984 - acc: 0.9724 - ETA: 2s - loss: 0.0977 - acc: 0.9725 - ETA: 2s - loss: 0.0977 - acc: 0.9724 - ETA: 2s - loss: 0.0966 - acc: 0.9726 - ETA: 2s - loss: 0.0973 - acc: 0.9720 - ETA: 2s - loss: 0.0968 - acc: 0.9722 - ETA: 2s - loss: 0.0959 - acc: 0.9724 - ETA: 2s - loss: 0.0950 - acc: 0.9725 - ETA: 2s - loss: 0.0959 - acc: 0.9717 - ETA: 2s - loss: 0.0962 - acc: 0.9714 - ETA: 2s - loss: 0.0950 - acc: 0.9718 - ETA: 2s - loss: 0.0945 - acc: 0.9717 - ETA: 2s - loss: 0.0936 - acc: 0.9719 - ETA: 2s - loss: 0.0927 - acc: 0.9723 - ETA: 2s - loss: 0.0920 - acc: 0.9726 - ETA: 2s - loss: 0.0927 - acc: 0.9721 - ETA: 1s - loss: 0.0937 - acc: 0.9716 - ETA: 1s - loss: 0.0929 - acc: 0.9718 - ETA: 1s - loss: 0.0941 - acc: 0.9713 - ETA: 1s - loss: 0.0933 - acc: 0.9717 - ETA: 1s - loss: 0.0957 - acc: 0.9714 - ETA: 1s - loss: 0.0950 - acc: 0.9717 - ETA: 1s - loss: 0.0948 - acc: 0.9719 - ETA: 1s - loss: 0.0956 - acc: 0.9716 - ETA: 1s - loss: 0.0947 - acc: 0.9719 - ETA: 1s - loss: 0.0952 - acc: 0.9719 - ETA: 1s - loss: 0.0948 - acc: 0.9716 - ETA: 1s - loss: 0.0941 - acc: 0.9718 - ETA: 1s - loss: 0.0935 - acc: 0.9719 - ETA: 1s - loss: 0.0937 - acc: 0.9717 - ETA: 1s - loss: 0.0944 - acc: 0.9714 - ETA: 1s - loss: 0.0937 - acc: 0.9717 - ETA: 0s - loss: 0.0944 - acc: 0.9710 - ETA: 0s - loss: 0.0941 - acc: 0.9711 - ETA: 0s - loss: 0.0961 - acc: 0.9705 - ETA: 0s - loss: 0.0967 - acc: 0.9702 - ETA: 0s - loss: 0.0975 - acc: 0.9698 - ETA: 0s - loss: 0.0966 - acc: 0.9701 - ETA: 0s - loss: 0.0960 - acc: 0.9704 - ETA: 0s - loss: 0.0982 - acc: 0.9702 - ETA: 0s - loss: 0.0996 - acc: 0.9697 - ETA: 0s - loss: 0.0990 - acc: 0.9698 - ETA: 0s - loss: 0.0984 - acc: 0.9701 - ETA: 0s - loss: 0.0976 - acc: 0.9704 - ETA: 0s - loss: 0.0971 - acc: 0.9705 - ETA: 0s - loss: 0.0986 - acc: 0.9705 - ETA: 0s - loss: 0.0985 - acc: 0.9704 - ETA: 0s - loss: 0.0985 - acc: 0.9702Epoch 00014: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0979 - acc: 0.9705 - val_loss: 0.6209 - val_acc: 0.8599
    Epoch 16/20
    6660/6680 [============================>.] - ETA: 6s - loss: 0.1650 - acc: 0.9500 - ETA: 6s - loss: 0.0590 - acc: 0.9875 - ETA: 6s - loss: 0.0965 - acc: 0.9786 - ETA: 6s - loss: 0.0832 - acc: 0.9750 - ETA: 6s - loss: 0.0791 - acc: 0.9769 - ETA: 6s - loss: 0.1104 - acc: 0.9719 - ETA: 6s - loss: 0.1056 - acc: 0.9737 - ETA: 6s - loss: 0.0971 - acc: 0.9750 - ETA: 6s - loss: 0.0941 - acc: 0.9760 - ETA: 6s - loss: 0.0888 - acc: 0.9768 - ETA: 6s - loss: 0.0815 - acc: 0.9790 - ETA: 6s - loss: 0.0804 - acc: 0.9779 - ETA: 6s - loss: 0.0767 - acc: 0.9784 - ETA: 6s - loss: 0.0774 - acc: 0.9775 - ETA: 6s - loss: 0.0730 - acc: 0.9791 - ETA: 6s - loss: 0.0783 - acc: 0.9793 - ETA: 6s - loss: 0.0788 - acc: 0.9796 - ETA: 6s - loss: 0.0754 - acc: 0.9808 - ETA: 5s - loss: 0.0805 - acc: 0.9782 - ETA: 5s - loss: 0.0793 - acc: 0.9784 - ETA: 5s - loss: 0.0818 - acc: 0.9770 - ETA: 5s - loss: 0.0806 - acc: 0.9766 - ETA: 5s - loss: 0.0833 - acc: 0.9761 - ETA: 5s - loss: 0.0816 - acc: 0.9771 - ETA: 5s - loss: 0.0863 - acc: 0.9767 - ETA: 5s - loss: 0.0840 - acc: 0.9770 - ETA: 5s - loss: 0.0843 - acc: 0.9766 - ETA: 5s - loss: 0.0851 - acc: 0.9762 - ETA: 5s - loss: 0.0881 - acc: 0.9759 - ETA: 5s - loss: 0.0871 - acc: 0.9761 - ETA: 5s - loss: 0.0899 - acc: 0.9742 - ETA: 5s - loss: 0.0894 - acc: 0.9734 - ETA: 5s - loss: 0.0876 - acc: 0.9737 - ETA: 5s - loss: 0.0865 - acc: 0.9742 - ETA: 4s - loss: 0.0860 - acc: 0.9745 - ETA: 4s - loss: 0.0924 - acc: 0.9743 - ETA: 4s - loss: 0.0911 - acc: 0.9750 - ETA: 4s - loss: 0.0892 - acc: 0.9757 - ETA: 4s - loss: 0.0881 - acc: 0.9754 - ETA: 4s - loss: 0.0924 - acc: 0.9744 - ETA: 4s - loss: 0.0910 - acc: 0.9746 - ETA: 4s - loss: 0.0916 - acc: 0.9748 - ETA: 4s - loss: 0.0908 - acc: 0.9746 - ETA: 4s - loss: 0.0889 - acc: 0.9752 - ETA: 4s - loss: 0.0879 - acc: 0.9750 - ETA: 4s - loss: 0.0894 - acc: 0.9752 - ETA: 4s - loss: 0.0880 - acc: 0.9754 - ETA: 4s - loss: 0.0876 - acc: 0.9755 - ETA: 4s - loss: 0.0882 - acc: 0.9750 - ETA: 3s - loss: 0.0894 - acc: 0.9741 - ETA: 3s - loss: 0.0889 - acc: 0.9743 - ETA: 3s - loss: 0.0891 - acc: 0.9739 - ETA: 3s - loss: 0.0884 - acc: 0.9744 - ETA: 3s - loss: 0.0874 - acc: 0.9745 - ETA: 3s - loss: 0.0866 - acc: 0.9747 - ETA: 3s - loss: 0.0860 - acc: 0.9748 - ETA: 3s - loss: 0.0854 - acc: 0.9750 - ETA: 3s - loss: 0.0857 - acc: 0.9751 - ETA: 3s - loss: 0.0848 - acc: 0.9756 - ETA: 3s - loss: 0.0858 - acc: 0.9751 - ETA: 3s - loss: 0.0845 - acc: 0.9756 - ETA: 3s - loss: 0.0843 - acc: 0.9754 - ETA: 3s - loss: 0.0841 - acc: 0.9753 - ETA: 3s - loss: 0.0829 - acc: 0.9757 - ETA: 2s - loss: 0.0827 - acc: 0.9755 - ETA: 2s - loss: 0.0826 - acc: 0.9754 - ETA: 2s - loss: 0.0818 - acc: 0.9755 - ETA: 2s - loss: 0.0810 - acc: 0.9756 - ETA: 2s - loss: 0.0842 - acc: 0.9755 - ETA: 2s - loss: 0.0834 - acc: 0.9754 - ETA: 2s - loss: 0.0853 - acc: 0.9752 - ETA: 2s - loss: 0.0844 - acc: 0.9756 - ETA: 2s - loss: 0.0841 - acc: 0.9755 - ETA: 2s - loss: 0.0850 - acc: 0.9753 - ETA: 2s - loss: 0.0848 - acc: 0.9752 - ETA: 2s - loss: 0.0838 - acc: 0.9756 - ETA: 2s - loss: 0.0831 - acc: 0.9757 - ETA: 2s - loss: 0.0835 - acc: 0.9753 - ETA: 2s - loss: 0.0834 - acc: 0.9754 - ETA: 2s - loss: 0.0845 - acc: 0.9749 - ETA: 1s - loss: 0.0840 - acc: 0.9750 - ETA: 1s - loss: 0.0839 - acc: 0.9751 - ETA: 1s - loss: 0.0847 - acc: 0.9748 - ETA: 1s - loss: 0.0852 - acc: 0.9743 - ETA: 1s - loss: 0.0866 - acc: 0.9740 - ETA: 1s - loss: 0.0870 - acc: 0.9739 - ETA: 1s - loss: 0.0870 - acc: 0.9738 - ETA: 1s - loss: 0.0865 - acc: 0.9739 - ETA: 1s - loss: 0.0875 - acc: 0.9733 - ETA: 1s - loss: 0.0874 - acc: 0.9734 - ETA: 1s - loss: 0.0885 - acc: 0.9730 - ETA: 1s - loss: 0.0879 - acc: 0.9733 - ETA: 1s - loss: 0.0877 - acc: 0.9734 - ETA: 1s - loss: 0.0869 - acc: 0.9737 - ETA: 1s - loss: 0.0863 - acc: 0.9739 - ETA: 1s - loss: 0.0858 - acc: 0.9740 - ETA: 0s - loss: 0.0862 - acc: 0.9740 - ETA: 0s - loss: 0.0858 - acc: 0.9741 - ETA: 0s - loss: 0.0864 - acc: 0.9738 - ETA: 0s - loss: 0.0875 - acc: 0.9736 - ETA: 0s - loss: 0.0872 - acc: 0.9737 - ETA: 0s - loss: 0.0876 - acc: 0.9736 - ETA: 0s - loss: 0.0873 - acc: 0.9735 - ETA: 0s - loss: 0.0885 - acc: 0.9733 - ETA: 0s - loss: 0.0908 - acc: 0.9728 - ETA: 0s - loss: 0.0918 - acc: 0.9727 - ETA: 0s - loss: 0.0916 - acc: 0.9726 - ETA: 0s - loss: 0.0912 - acc: 0.9727 - ETA: 0s - loss: 0.0909 - acc: 0.9728 - ETA: 0s - loss: 0.0902 - acc: 0.9731 - ETA: 0s - loss: 0.0916 - acc: 0.9730 - ETA: 0s - loss: 0.0923 - acc: 0.9728Epoch 00015: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0921 - acc: 0.9729 - val_loss: 0.6234 - val_acc: 0.8635
    Epoch 17/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.2965 - acc: 0.9500 - ETA: 6s - loss: 0.1345 - acc: 0.9625 - ETA: 6s - loss: 0.1052 - acc: 0.9643 - ETA: 6s - loss: 0.0859 - acc: 0.9650 - ETA: 6s - loss: 0.0742 - acc: 0.9654 - ETA: 6s - loss: 0.0642 - acc: 0.9687 - ETA: 6s - loss: 0.0588 - acc: 0.9737 - ETA: 6s - loss: 0.0668 - acc: 0.9727 - ETA: 6s - loss: 0.0614 - acc: 0.9760 - ETA: 6s - loss: 0.0608 - acc: 0.9768 - ETA: 6s - loss: 0.0556 - acc: 0.9790 - ETA: 6s - loss: 0.0528 - acc: 0.9809 - ETA: 6s - loss: 0.0501 - acc: 0.9824 - ETA: 6s - loss: 0.0485 - acc: 0.9825 - ETA: 6s - loss: 0.0619 - acc: 0.9779 - ETA: 6s - loss: 0.0713 - acc: 0.9761 - ETA: 6s - loss: 0.0750 - acc: 0.9755 - ETA: 5s - loss: 0.0737 - acc: 0.9750 - ETA: 5s - loss: 0.0759 - acc: 0.9745 - ETA: 5s - loss: 0.0837 - acc: 0.9741 - ETA: 5s - loss: 0.0806 - acc: 0.9754 - ETA: 5s - loss: 0.0795 - acc: 0.9750 - ETA: 5s - loss: 0.0789 - acc: 0.9754 - ETA: 5s - loss: 0.0778 - acc: 0.9757 - ETA: 5s - loss: 0.0849 - acc: 0.9747 - ETA: 5s - loss: 0.0851 - acc: 0.9750 - ETA: 5s - loss: 0.0825 - acc: 0.9759 - ETA: 5s - loss: 0.0798 - acc: 0.9768 - ETA: 5s - loss: 0.0790 - acc: 0.9771 - ETA: 5s - loss: 0.0795 - acc: 0.9773 - ETA: 5s - loss: 0.0782 - acc: 0.9775 - ETA: 5s - loss: 0.0764 - acc: 0.9777 - ETA: 5s - loss: 0.0822 - acc: 0.9768 - ETA: 4s - loss: 0.0843 - acc: 0.9755 - ETA: 4s - loss: 0.0931 - acc: 0.9743 - ETA: 4s - loss: 0.0908 - acc: 0.9750 - ETA: 4s - loss: 0.0891 - acc: 0.9757 - ETA: 4s - loss: 0.0876 - acc: 0.9759 - ETA: 4s - loss: 0.0870 - acc: 0.9757 - ETA: 4s - loss: 0.0855 - acc: 0.9763 - ETA: 4s - loss: 0.0846 - acc: 0.9764 - ETA: 4s - loss: 0.0836 - acc: 0.9762 - ETA: 4s - loss: 0.0853 - acc: 0.9752 - ETA: 4s - loss: 0.0841 - acc: 0.9754 - ETA: 4s - loss: 0.0839 - acc: 0.9748 - ETA: 4s - loss: 0.0835 - acc: 0.9750 - ETA: 4s - loss: 0.0837 - acc: 0.9748 - ETA: 4s - loss: 0.0824 - acc: 0.9754 - ETA: 4s - loss: 0.0846 - acc: 0.9748 - ETA: 3s - loss: 0.0861 - acc: 0.9750 - ETA: 3s - loss: 0.0861 - acc: 0.9748 - ETA: 3s - loss: 0.0855 - acc: 0.9750 - ETA: 3s - loss: 0.0877 - acc: 0.9742 - ETA: 3s - loss: 0.0891 - acc: 0.9737 - ETA: 3s - loss: 0.0879 - acc: 0.9739 - ETA: 3s - loss: 0.0875 - acc: 0.9741 - ETA: 3s - loss: 0.0863 - acc: 0.9746 - ETA: 3s - loss: 0.0860 - acc: 0.9747 - ETA: 3s - loss: 0.0860 - acc: 0.9743 - ETA: 3s - loss: 0.0856 - acc: 0.9744 - ETA: 3s - loss: 0.0845 - acc: 0.9749 - ETA: 3s - loss: 0.0834 - acc: 0.9753 - ETA: 3s - loss: 0.0852 - acc: 0.9746 - ETA: 3s - loss: 0.0840 - acc: 0.9750 - ETA: 3s - loss: 0.0829 - acc: 0.9754 - ETA: 2s - loss: 0.0821 - acc: 0.9755 - ETA: 2s - loss: 0.0809 - acc: 0.9759 - ETA: 2s - loss: 0.0824 - acc: 0.9757 - ETA: 2s - loss: 0.0858 - acc: 0.9751 - ETA: 2s - loss: 0.0855 - acc: 0.9748 - ETA: 2s - loss: 0.0852 - acc: 0.9746 - ETA: 2s - loss: 0.0858 - acc: 0.9743 - ETA: 2s - loss: 0.0850 - acc: 0.9744 - ETA: 2s - loss: 0.0855 - acc: 0.9741 - ETA: 2s - loss: 0.0866 - acc: 0.9738 - ETA: 2s - loss: 0.0868 - acc: 0.9739 - ETA: 2s - loss: 0.0869 - acc: 0.9738 - ETA: 2s - loss: 0.0915 - acc: 0.9733 - ETA: 2s - loss: 0.0923 - acc: 0.9730 - ETA: 2s - loss: 0.0924 - acc: 0.9729 - ETA: 1s - loss: 0.0921 - acc: 0.9730 - ETA: 1s - loss: 0.0918 - acc: 0.9732 - ETA: 1s - loss: 0.0910 - acc: 0.9733 - ETA: 1s - loss: 0.0931 - acc: 0.9726 - ETA: 1s - loss: 0.0926 - acc: 0.9725 - ETA: 1s - loss: 0.0931 - acc: 0.9725 - ETA: 1s - loss: 0.0926 - acc: 0.9726 - ETA: 1s - loss: 0.0927 - acc: 0.9725 - ETA: 1s - loss: 0.0955 - acc: 0.9721 - ETA: 1s - loss: 0.0953 - acc: 0.9718 - ETA: 1s - loss: 0.0945 - acc: 0.9721 - ETA: 1s - loss: 0.0939 - acc: 0.9723 - ETA: 1s - loss: 0.0932 - acc: 0.9722 - ETA: 1s - loss: 0.0938 - acc: 0.9720 - ETA: 1s - loss: 0.0933 - acc: 0.9721 - ETA: 1s - loss: 0.0924 - acc: 0.9724 - ETA: 0s - loss: 0.0928 - acc: 0.9723 - ETA: 0s - loss: 0.0919 - acc: 0.9726 - ETA: 0s - loss: 0.0921 - acc: 0.9725 - ETA: 0s - loss: 0.0922 - acc: 0.9723 - ETA: 0s - loss: 0.0914 - acc: 0.9726 - ETA: 0s - loss: 0.0925 - acc: 0.9722 - ETA: 0s - loss: 0.0921 - acc: 0.9721 - ETA: 0s - loss: 0.0915 - acc: 0.9724 - ETA: 0s - loss: 0.0919 - acc: 0.9724 - ETA: 0s - loss: 0.0913 - acc: 0.9726 - ETA: 0s - loss: 0.0906 - acc: 0.9727 - ETA: 0s - loss: 0.0908 - acc: 0.9727 - ETA: 0s - loss: 0.0901 - acc: 0.9728 - ETA: 0s - loss: 0.0899 - acc: 0.9729 - ETA: 0s - loss: 0.0896 - acc: 0.9730Epoch 00016: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0898 - acc: 0.9731 - val_loss: 0.6858 - val_acc: 0.8455
    Epoch 18/20
    6660/6680 [============================>.] - ETA: 6s - loss: 0.0013 - acc: 1.0000 - ETA: 6s - loss: 0.0212 - acc: 1.0000 - ETA: 6s - loss: 0.0408 - acc: 0.9857 - ETA: 6s - loss: 0.0702 - acc: 0.9800 - ETA: 6s - loss: 0.0673 - acc: 0.9769 - ETA: 6s - loss: 0.0589 - acc: 0.9812 - ETA: 6s - loss: 0.0584 - acc: 0.9816 - ETA: 6s - loss: 0.0543 - acc: 0.9818 - ETA: 6s - loss: 0.0493 - acc: 0.9840 - ETA: 6s - loss: 0.0462 - acc: 0.9857 - ETA: 6s - loss: 0.0502 - acc: 0.9855 - ETA: 6s - loss: 0.0468 - acc: 0.9868 - ETA: 5s - loss: 0.0458 - acc: 0.9865 - ETA: 5s - loss: 0.0432 - acc: 0.9875 - ETA: 5s - loss: 0.0437 - acc: 0.9872 - ETA: 5s - loss: 0.0429 - acc: 0.9870 - ETA: 5s - loss: 0.0432 - acc: 0.9867 - ETA: 5s - loss: 0.0447 - acc: 0.9865 - ETA: 5s - loss: 0.0459 - acc: 0.9864 - ETA: 5s - loss: 0.0485 - acc: 0.9862 - ETA: 5s - loss: 0.0491 - acc: 0.9852 - ETA: 5s - loss: 0.0518 - acc: 0.9852 - ETA: 5s - loss: 0.0560 - acc: 0.9828 - ETA: 5s - loss: 0.0548 - acc: 0.9836 - ETA: 5s - loss: 0.0540 - acc: 0.9836 - ETA: 5s - loss: 0.0529 - acc: 0.9836 - ETA: 5s - loss: 0.0541 - acc: 0.9835 - ETA: 5s - loss: 0.0542 - acc: 0.9829 - ETA: 5s - loss: 0.0548 - acc: 0.9818 - ETA: 4s - loss: 0.0573 - acc: 0.9818 - ETA: 4s - loss: 0.0581 - acc: 0.9813 - ETA: 4s - loss: 0.0565 - acc: 0.9819 - ETA: 4s - loss: 0.0556 - acc: 0.9825 - ETA: 4s - loss: 0.0554 - acc: 0.9825 - ETA: 4s - loss: 0.0563 - acc: 0.9825 - ETA: 4s - loss: 0.0599 - acc: 0.9821 - ETA: 4s - loss: 0.0602 - acc: 0.9821 - ETA: 4s - loss: 0.0604 - acc: 0.9817 - ETA: 4s - loss: 0.0600 - acc: 0.9817 - ETA: 4s - loss: 0.0603 - acc: 0.9814 - ETA: 4s - loss: 0.0622 - acc: 0.9806 - ETA: 4s - loss: 0.0660 - acc: 0.9806 - ETA: 4s - loss: 0.0691 - acc: 0.9799 - ETA: 4s - loss: 0.0690 - acc: 0.9800 - ETA: 4s - loss: 0.0706 - acc: 0.9801 - ETA: 3s - loss: 0.0724 - acc: 0.9798 - ETA: 3s - loss: 0.0724 - acc: 0.9799 - ETA: 3s - loss: 0.0741 - acc: 0.9796 - ETA: 3s - loss: 0.0739 - acc: 0.9797 - ETA: 3s - loss: 0.0729 - acc: 0.9801 - ETA: 3s - loss: 0.0732 - acc: 0.9798 - ETA: 3s - loss: 0.0752 - acc: 0.9786 - ETA: 3s - loss: 0.0744 - acc: 0.9787 - ETA: 3s - loss: 0.0735 - acc: 0.9791 - ETA: 3s - loss: 0.0723 - acc: 0.9794 - ETA: 3s - loss: 0.0729 - acc: 0.9789 - ETA: 3s - loss: 0.0721 - acc: 0.9790 - ETA: 3s - loss: 0.0715 - acc: 0.9791 - ETA: 3s - loss: 0.0727 - acc: 0.9791 - ETA: 3s - loss: 0.0742 - acc: 0.9792 - ETA: 3s - loss: 0.0731 - acc: 0.9796 - ETA: 3s - loss: 0.0723 - acc: 0.9796 - ETA: 2s - loss: 0.0716 - acc: 0.9797 - ETA: 2s - loss: 0.0712 - acc: 0.9797 - ETA: 2s - loss: 0.0704 - acc: 0.9801 - ETA: 2s - loss: 0.0720 - acc: 0.9798 - ETA: 2s - loss: 0.0724 - acc: 0.9796 - ETA: 2s - loss: 0.0734 - acc: 0.9790 - ETA: 2s - loss: 0.0731 - acc: 0.9788 - ETA: 2s - loss: 0.0724 - acc: 0.9788 - ETA: 2s - loss: 0.0734 - acc: 0.9784 - ETA: 2s - loss: 0.0729 - acc: 0.9783 - ETA: 2s - loss: 0.0720 - acc: 0.9786 - ETA: 2s - loss: 0.0718 - acc: 0.9786 - ETA: 2s - loss: 0.0751 - acc: 0.9780 - ETA: 2s - loss: 0.0756 - acc: 0.9774 - ETA: 2s - loss: 0.0769 - acc: 0.9773 - ETA: 2s - loss: 0.0762 - acc: 0.9776 - ETA: 2s - loss: 0.0766 - acc: 0.9774 - ETA: 1s - loss: 0.0761 - acc: 0.9775 - ETA: 1s - loss: 0.0756 - acc: 0.9776 - ETA: 1s - loss: 0.0749 - acc: 0.9779 - ETA: 1s - loss: 0.0744 - acc: 0.9779 - ETA: 1s - loss: 0.0767 - acc: 0.9774 - ETA: 1s - loss: 0.0774 - acc: 0.9769 - ETA: 1s - loss: 0.0776 - acc: 0.9768 - ETA: 1s - loss: 0.0781 - acc: 0.9763 - ETA: 1s - loss: 0.0774 - acc: 0.9765 - ETA: 1s - loss: 0.0768 - acc: 0.9768 - ETA: 1s - loss: 0.0817 - acc: 0.9761 - ETA: 1s - loss: 0.0813 - acc: 0.9762 - ETA: 1s - loss: 0.0844 - acc: 0.9755 - ETA: 1s - loss: 0.0845 - acc: 0.9755 - ETA: 1s - loss: 0.0840 - acc: 0.9757 - ETA: 1s - loss: 0.0846 - acc: 0.9754 - ETA: 1s - loss: 0.0847 - acc: 0.9755 - ETA: 0s - loss: 0.0841 - acc: 0.9758 - ETA: 0s - loss: 0.0835 - acc: 0.9760 - ETA: 0s - loss: 0.0828 - acc: 0.9763 - ETA: 0s - loss: 0.0821 - acc: 0.9765 - ETA: 0s - loss: 0.0816 - acc: 0.9768 - ETA: 0s - loss: 0.0812 - acc: 0.9768 - ETA: 0s - loss: 0.0827 - acc: 0.9769 - ETA: 0s - loss: 0.0821 - acc: 0.9771 - ETA: 0s - loss: 0.0834 - acc: 0.9767 - ETA: 0s - loss: 0.0830 - acc: 0.9768 - ETA: 0s - loss: 0.0834 - acc: 0.9763 - ETA: 0s - loss: 0.0830 - acc: 0.9762 - ETA: 0s - loss: 0.0830 - acc: 0.9762 - ETA: 0s - loss: 0.0831 - acc: 0.9762 - ETA: 0s - loss: 0.0833 - acc: 0.9760 - ETA: 0s - loss: 0.0830 - acc: 0.9761 - ETA: 0s - loss: 0.0828 - acc: 0.9761 - ETA: 0s - loss: 0.0825 - acc: 0.9761Epoch 00017: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0823 - acc: 0.9762 - val_loss: 0.6616 - val_acc: 0.8527
    Epoch 19/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.0019 - acc: 1.0000 - ETA: 6s - loss: 0.0174 - acc: 0.9875 - ETA: 6s - loss: 0.0192 - acc: 0.9929 - ETA: 6s - loss: 0.0262 - acc: 0.9900 - ETA: 6s - loss: 0.0610 - acc: 0.9731 - ETA: 6s - loss: 0.0690 - acc: 0.9750 - ETA: 6s - loss: 0.0730 - acc: 0.9711 - ETA: 6s - loss: 0.0747 - acc: 0.9705 - ETA: 6s - loss: 0.0776 - acc: 0.9720 - ETA: 6s - loss: 0.0742 - acc: 0.9732 - ETA: 6s - loss: 0.0690 - acc: 0.9742 - ETA: 6s - loss: 0.0634 - acc: 0.9765 - ETA: 6s - loss: 0.0733 - acc: 0.9743 - ETA: 6s - loss: 0.0706 - acc: 0.9750 - ETA: 6s - loss: 0.0672 - acc: 0.9767 - ETA: 6s - loss: 0.0702 - acc: 0.9772 - ETA: 6s - loss: 0.0670 - acc: 0.9786 - ETA: 6s - loss: 0.0700 - acc: 0.9769 - ETA: 5s - loss: 0.0689 - acc: 0.9773 - ETA: 5s - loss: 0.0664 - acc: 0.9784 - ETA: 5s - loss: 0.0640 - acc: 0.9795 - ETA: 5s - loss: 0.0616 - acc: 0.9805 - ETA: 5s - loss: 0.0622 - acc: 0.9799 - ETA: 5s - loss: 0.0607 - acc: 0.9800 - ETA: 5s - loss: 0.0622 - acc: 0.9788 - ETA: 5s - loss: 0.0642 - acc: 0.9783 - ETA: 5s - loss: 0.0625 - acc: 0.9785 - ETA: 5s - loss: 0.0677 - acc: 0.9774 - ETA: 5s - loss: 0.0732 - acc: 0.9771 - ETA: 5s - loss: 0.0740 - acc: 0.9773 - ETA: 5s - loss: 0.0742 - acc: 0.9769 - ETA: 5s - loss: 0.0721 - acc: 0.9777 - ETA: 5s - loss: 0.0718 - acc: 0.9773 - ETA: 5s - loss: 0.0728 - acc: 0.9765 - ETA: 4s - loss: 0.0713 - acc: 0.9767 - ETA: 4s - loss: 0.0695 - acc: 0.9774 - ETA: 4s - loss: 0.0720 - acc: 0.9775 - ETA: 4s - loss: 0.0704 - acc: 0.9781 - ETA: 4s - loss: 0.0699 - acc: 0.9778 - ETA: 4s - loss: 0.0709 - acc: 0.9775 - ETA: 4s - loss: 0.0710 - acc: 0.9773 - ETA: 4s - loss: 0.0696 - acc: 0.9778 - ETA: 4s - loss: 0.0730 - acc: 0.9768 - ETA: 4s - loss: 0.0733 - acc: 0.9762 - ETA: 4s - loss: 0.0723 - acc: 0.9767 - ETA: 4s - loss: 0.0721 - acc: 0.9765 - ETA: 4s - loss: 0.0742 - acc: 0.9763 - ETA: 4s - loss: 0.0731 - acc: 0.9768 - ETA: 4s - loss: 0.0734 - acc: 0.9766 - ETA: 3s - loss: 0.0728 - acc: 0.9767 - ETA: 3s - loss: 0.0722 - acc: 0.9768 - ETA: 3s - loss: 0.0724 - acc: 0.9769 - ETA: 3s - loss: 0.0718 - acc: 0.9771 - ETA: 3s - loss: 0.0765 - acc: 0.9769 - ETA: 3s - loss: 0.0755 - acc: 0.9770 - ETA: 3s - loss: 0.0754 - acc: 0.9771 - ETA: 3s - loss: 0.0743 - acc: 0.9775 - ETA: 3s - loss: 0.0785 - acc: 0.9773 - ETA: 3s - loss: 0.0775 - acc: 0.9777 - ETA: 3s - loss: 0.0766 - acc: 0.9778 - ETA: 3s - loss: 0.0791 - acc: 0.9776 - ETA: 3s - loss: 0.0791 - acc: 0.9777 - ETA: 3s - loss: 0.0790 - acc: 0.9775 - ETA: 3s - loss: 0.0799 - acc: 0.9774 - ETA: 3s - loss: 0.0798 - acc: 0.9775 - ETA: 2s - loss: 0.0800 - acc: 0.9770 - ETA: 2s - loss: 0.0797 - acc: 0.9771 - ETA: 2s - loss: 0.0790 - acc: 0.9775 - ETA: 2s - loss: 0.0801 - acc: 0.9773 - ETA: 2s - loss: 0.0792 - acc: 0.9776 - ETA: 2s - loss: 0.0794 - acc: 0.9770 - ETA: 2s - loss: 0.0784 - acc: 0.9773 - ETA: 2s - loss: 0.0786 - acc: 0.9772 - ETA: 2s - loss: 0.0777 - acc: 0.9775 - ETA: 2s - loss: 0.0770 - acc: 0.9778 - ETA: 2s - loss: 0.0766 - acc: 0.9779 - ETA: 2s - loss: 0.0764 - acc: 0.9777 - ETA: 2s - loss: 0.0773 - acc: 0.9778 - ETA: 2s - loss: 0.0771 - acc: 0.9779 - ETA: 2s - loss: 0.0763 - acc: 0.9782 - ETA: 1s - loss: 0.0780 - acc: 0.9778 - ETA: 1s - loss: 0.0775 - acc: 0.9779 - ETA: 1s - loss: 0.0779 - acc: 0.9775 - ETA: 1s - loss: 0.0771 - acc: 0.9778 - ETA: 1s - loss: 0.0767 - acc: 0.9779 - ETA: 1s - loss: 0.0758 - acc: 0.9781 - ETA: 1s - loss: 0.0755 - acc: 0.9780 - ETA: 1s - loss: 0.0747 - acc: 0.9782 - ETA: 1s - loss: 0.0752 - acc: 0.9781 - ETA: 1s - loss: 0.0752 - acc: 0.9780 - ETA: 1s - loss: 0.0745 - acc: 0.9782 - ETA: 1s - loss: 0.0743 - acc: 0.9781 - ETA: 1s - loss: 0.0742 - acc: 0.9780 - ETA: 1s - loss: 0.0752 - acc: 0.9779 - ETA: 1s - loss: 0.0746 - acc: 0.9781 - ETA: 1s - loss: 0.0751 - acc: 0.9778 - ETA: 0s - loss: 0.0754 - acc: 0.9779 - ETA: 0s - loss: 0.0763 - acc: 0.9776 - ETA: 0s - loss: 0.0762 - acc: 0.9773 - ETA: 0s - loss: 0.0769 - acc: 0.9770 - ETA: 0s - loss: 0.0765 - acc: 0.9771 - ETA: 0s - loss: 0.0761 - acc: 0.9771 - ETA: 0s - loss: 0.0755 - acc: 0.9774 - ETA: 0s - loss: 0.0766 - acc: 0.9771 - ETA: 0s - loss: 0.0762 - acc: 0.9773 - ETA: 0s - loss: 0.0763 - acc: 0.9771 - ETA: 0s - loss: 0.0781 - acc: 0.9766 - ETA: 0s - loss: 0.0774 - acc: 0.9769 - ETA: 0s - loss: 0.0780 - acc: 0.9769 - ETA: 0s - loss: 0.0774 - acc: 0.9771 - ETA: 0s - loss: 0.0771 - acc: 0.9772Epoch 00018: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0771 - acc: 0.9772 - val_loss: 0.6721 - val_acc: 0.8575
    Epoch 20/20
    6620/6680 [============================>.] - ETA: 6s - loss: 0.1905 - acc: 0.8500 - ETA: 6s - loss: 0.0562 - acc: 0.9625 - ETA: 6s - loss: 0.0356 - acc: 0.9786 - ETA: 6s - loss: 0.0276 - acc: 0.9850 - ETA: 6s - loss: 0.0380 - acc: 0.9846 - ETA: 6s - loss: 0.0459 - acc: 0.9781 - ETA: 6s - loss: 0.0394 - acc: 0.9816 - ETA: 6s - loss: 0.0405 - acc: 0.9818 - ETA: 6s - loss: 0.0559 - acc: 0.9800 - ETA: 6s - loss: 0.0733 - acc: 0.9786 - ETA: 6s - loss: 0.0707 - acc: 0.9790 - ETA: 6s - loss: 0.0657 - acc: 0.9809 - ETA: 5s - loss: 0.0625 - acc: 0.9824 - ETA: 5s - loss: 0.0598 - acc: 0.9837 - ETA: 5s - loss: 0.0567 - acc: 0.9849 - ETA: 5s - loss: 0.0669 - acc: 0.9837 - ETA: 5s - loss: 0.0645 - acc: 0.9847 - ETA: 5s - loss: 0.0616 - acc: 0.9856 - ETA: 5s - loss: 0.0609 - acc: 0.9855 - ETA: 5s - loss: 0.0590 - acc: 0.9853 - ETA: 5s - loss: 0.0590 - acc: 0.9852 - ETA: 5s - loss: 0.0579 - acc: 0.9852 - ETA: 5s - loss: 0.0567 - acc: 0.9851 - ETA: 5s - loss: 0.0548 - acc: 0.9857 - ETA: 5s - loss: 0.0550 - acc: 0.9849 - ETA: 5s - loss: 0.0540 - acc: 0.9849 - ETA: 5s - loss: 0.0536 - acc: 0.9848 - ETA: 5s - loss: 0.0519 - acc: 0.9854 - ETA: 5s - loss: 0.0545 - acc: 0.9841 - ETA: 4s - loss: 0.0533 - acc: 0.9847 - ETA: 4s - loss: 0.0521 - acc: 0.9852 - ETA: 4s - loss: 0.0528 - acc: 0.9851 - ETA: 4s - loss: 0.0537 - acc: 0.9845 - ETA: 4s - loss: 0.0532 - acc: 0.9845 - ETA: 4s - loss: 0.0557 - acc: 0.9835 - ETA: 4s - loss: 0.0545 - acc: 0.9840 - ETA: 4s - loss: 0.0563 - acc: 0.9835 - ETA: 4s - loss: 0.0554 - acc: 0.9835 - ETA: 4s - loss: 0.0611 - acc: 0.9822 - ETA: 4s - loss: 0.0613 - acc: 0.9818 - ETA: 4s - loss: 0.0600 - acc: 0.9822 - ETA: 4s - loss: 0.0607 - acc: 0.9819 - ETA: 4s - loss: 0.0619 - acc: 0.9811 - ETA: 4s - loss: 0.0615 - acc: 0.9808 - ETA: 4s - loss: 0.0606 - acc: 0.9812 - ETA: 3s - loss: 0.0598 - acc: 0.9816 - ETA: 3s - loss: 0.0607 - acc: 0.9817 - ETA: 3s - loss: 0.0621 - acc: 0.9813 - ETA: 3s - loss: 0.0619 - acc: 0.9814 - ETA: 3s - loss: 0.0612 - acc: 0.9818 - ETA: 3s - loss: 0.0612 - acc: 0.9815 - ETA: 3s - loss: 0.0616 - acc: 0.9808 - ETA: 3s - loss: 0.0640 - acc: 0.9796 - ETA: 3s - loss: 0.0652 - acc: 0.9787 - ETA: 3s - loss: 0.0655 - acc: 0.9785 - ETA: 3s - loss: 0.0653 - acc: 0.9783 - ETA: 3s - loss: 0.0656 - acc: 0.9781 - ETA: 3s - loss: 0.0669 - acc: 0.9782 - ETA: 3s - loss: 0.0680 - acc: 0.9780 - ETA: 3s - loss: 0.0714 - acc: 0.9770 - ETA: 3s - loss: 0.0711 - acc: 0.9768 - ETA: 3s - loss: 0.0700 - acc: 0.9772 - ETA: 2s - loss: 0.0705 - acc: 0.9770 - ETA: 2s - loss: 0.0714 - acc: 0.9768 - ETA: 2s - loss: 0.0724 - acc: 0.9769 - ETA: 2s - loss: 0.0731 - acc: 0.9768 - ETA: 2s - loss: 0.0727 - acc: 0.9766 - ETA: 2s - loss: 0.0717 - acc: 0.9770 - ETA: 2s - loss: 0.0718 - acc: 0.9771 - ETA: 2s - loss: 0.0755 - acc: 0.9769 - ETA: 2s - loss: 0.0756 - acc: 0.9768 - ETA: 2s - loss: 0.0749 - acc: 0.9771 - ETA: 2s - loss: 0.0756 - acc: 0.9772 - ETA: 2s - loss: 0.0751 - acc: 0.9775 - ETA: 2s - loss: 0.0746 - acc: 0.9774 - ETA: 2s - loss: 0.0747 - acc: 0.9772 - ETA: 2s - loss: 0.0749 - acc: 0.9773 - ETA: 2s - loss: 0.0765 - acc: 0.9769 - ETA: 2s - loss: 0.0795 - acc: 0.9764 - ETA: 1s - loss: 0.0788 - acc: 0.9767 - ETA: 1s - loss: 0.0780 - acc: 0.9770 - ETA: 1s - loss: 0.0772 - acc: 0.9773 - ETA: 1s - loss: 0.0768 - acc: 0.9773 - ETA: 1s - loss: 0.0767 - acc: 0.9774 - ETA: 1s - loss: 0.0762 - acc: 0.9775 - ETA: 1s - loss: 0.0762 - acc: 0.9771 - ETA: 1s - loss: 0.0758 - acc: 0.9772 - ETA: 1s - loss: 0.0752 - acc: 0.9775 - ETA: 1s - loss: 0.0747 - acc: 0.9775 - ETA: 1s - loss: 0.0750 - acc: 0.9772 - ETA: 1s - loss: 0.0743 - acc: 0.9775 - ETA: 1s - loss: 0.0738 - acc: 0.9777 - ETA: 1s - loss: 0.0737 - acc: 0.9778 - ETA: 1s - loss: 0.0736 - acc: 0.9777 - ETA: 1s - loss: 0.0734 - acc: 0.9776 - ETA: 0s - loss: 0.0747 - acc: 0.9773 - ETA: 0s - loss: 0.0744 - acc: 0.9773 - ETA: 0s - loss: 0.0738 - acc: 0.9776 - ETA: 0s - loss: 0.0756 - acc: 0.9775 - ETA: 0s - loss: 0.0751 - acc: 0.9775 - ETA: 0s - loss: 0.0745 - acc: 0.9777 - ETA: 0s - loss: 0.0738 - acc: 0.9780 - ETA: 0s - loss: 0.0742 - acc: 0.9779 - ETA: 0s - loss: 0.0738 - acc: 0.9781 - ETA: 0s - loss: 0.0733 - acc: 0.9781 - ETA: 0s - loss: 0.0731 - acc: 0.9782 - ETA: 0s - loss: 0.0727 - acc: 0.9782 - ETA: 0s - loss: 0.0723 - acc: 0.9784 - ETA: 0s - loss: 0.0728 - acc: 0.9783 - ETA: 0s - loss: 0.0743 - acc: 0.9782 - ETA: 0s - loss: 0.0738 - acc: 0.9784Epoch 00019: val_loss did not improve
    6680/6680 [==============================] - 7s - loss: 0.0734 - acc: 0.9784 - val_loss: 0.7048 - val_acc: 0.8599
    ---I am done saving model valid_Xception  ----
    

    (IMPLEMENTATION) Load the Model with the Best Validation Loss

    In [42]:
    ### TODO: Load the model weights with the best validation loss.
    # valid_Xception loss: 0.0713 - acc: 0.9798 - val_loss: 0.6841 - val_acc: 0.8503 'weights.best.Xception.hdf5'
    # valid_InceptionV3 loss: 0.0302 - acc: 0.9906 - val_loss: 0.9732 - val_acc: 0.8383 'weights.best.InceptionV3.hdf5'
    # valid_Resnet50 loss: 0.0065 - acc: 0.9988 - val_loss: 0.9301 - val_acc: 0.8323 'weights.best.Resnet50.hdf5'
    # VGG1          loss: 7.5263 - acc: 0.5213 - val_loss: 8.1408 - val_acc: 0.4311 'weights.best.VGG19.hdf5'
    VGG19_model.load_weights('weights.best.VGG19.hdf5')
    print('-- VGG19 Weights Loaded --- ')
    Xception_model.load_weights('weights.best.Xception.hdf5')
    print('-- Xception Weights Loaded --- ')
    InceptionV3_model.load_weights('weights.best.InceptionV3.hdf5')
    print('-- InceptionV3 Weights Loaded --- ')
    Resnet50_model.load_weights('weights.best.Resnet50.hdf5')
    print('-- Resnet50 Weights Loaded --- ')
    
    -- VGG19 Weights Loaded --- 
    -- Xception Weights Loaded --- 
    -- InceptionV3 Weights Loaded --- 
    -- Resnet50 Weights Loaded --- 
    

    (IMPLEMENTATION) Test the Model

    Try out your model on the test dataset of dog images. Ensure that your test accuracy is greater than 60%.

    In [44]:
    ### TODO: Calculate classification accuracy on the test dataset.
    VGG19_predictions = [np.argmax(VGG19_model.predict(np.expand_dims(feature, axis=0))) for feature in test_VGG19]
    print('-- VGG19_predictions Loaded --- ')
    Xception_predictions = [np.argmax(Xception_model.predict(np.expand_dims(feature, axis=0))) for feature in test_Xception]
    print('-- Xception_predictions Loaded --- ')
    InceptionV3_predictions = [np.argmax(InceptionV3_model.predict(np.expand_dims(feature, axis=0))) for feature in test_InceptionV3]
    print('-- InceptionV3_predictions Loaded --- ')
    Resnet50_predictions = [np.argmax(Resnet50_model.predict(np.expand_dims(feature, axis=0))) for feature in test_Resnet50]
    print('-- Resnet50_predictions Loaded --- ')
    
    
    test_accuracy_VGG19 = 100*np.sum(np.array(VGG19_predictions)==np.argmax(test_targets, axis=1))/len(VGG19_predictions)
    print('VGG19 Test accuracy: %.4f%%' % test_accuracy_VGG19)
    
    test_accuracy_Xception = 100*np.sum(np.array(Xception_predictions)==np.argmax(test_targets, axis=1))/len(Xception_predictions)
    print('Xception Test accuracy: %.4f%%' % test_accuracy_Xception)
    
    test_accuracy_InceptionV3 = 100*np.sum(np.array(InceptionV3_predictions)==np.argmax(test_targets, axis=1))/len(InceptionV3_predictions)
    print('Inception Test accuracy: %.4f%%' % test_accuracy_InceptionV3)
    
    test_accuracy_Resnet50 = 100*np.sum(np.array(Resnet50_predictions)==np.argmax(test_targets, axis=1))/len(Resnet50_predictions)
    print('Resnet50 Test accuracy: %.4f%%' % test_accuracy_Resnet50)
    
    -- VGG19_predictions Loaded --- 
    -- Xception_predictions Loaded --- 
    -- InceptionV3_predictions Loaded --- 
    -- Resnet50_predictions Loaded --- 
    VGG19 Test accuracy: 49.7608%
    Xception Test accuracy: 85.0478%
    Inception Test accuracy: 77.1531%
    Resnet50 Test accuracy: 81.2201%
    

    Xception , Inception and Resnet provides Test accuracy higher than 60%
    Xception Test accuracy: 85.0478%
    Inception Test accuracy: 77.1531%
    Resnet50 Test accuracy: 81.2201%

    (IMPLEMENTATION) Predict Dog Breed with the Model

    Write a function that takes an image path as input and returns the dog breed (Affenpinscher, Afghan_hound, etc) that is predicted by your model.

    Similar to the analogous function in Step 5, your function should have three steps:

    1. Extract the bottleneck features corresponding to the chosen CNN model.
    2. Supply the bottleneck features as input to the model to return the predicted vector. Note that the argmax of this prediction vector gives the index of the predicted dog breed.
    3. Use the dog_names array defined in Step 0 of this notebook to return the corresponding breed.

    The functions to extract the bottleneck features can be found in extract_bottleneck_features.py, and they have been imported in an earlier code cell. To obtain the bottleneck features corresponding to your chosen CNN architecture, you need to use the function

    extract_{network}
    
    

    where {network}, in the above filename, should be one of VGG19, Resnet50, InceptionV3, or Xception.

    In [45]:
    ### TODO: Write a function that takes a path to an image as input
    ### and returns the dog breed that is predicted by the model.
    
    
    def predict_breed_for_dog(img_path):
        # extract bottleneck features
        bottleneck_feature = extract_Xception(path_to_tensor(img_path))
        # obtain predicted vector
        predicted_vector = Xception_model.predict(bottleneck_feature)
        # return dog breed that is predicted by the model
        return dog_names[np.argmax(predicted_vector)]
    
    print(' I am done with predict_breed_for_dog ')
    
     I am done with predict_breed_for_dog 
    

    Step 6: Write your Algorithm

    Write an algorithm that accepts a file path to an image and first determines whether the image contains a human, dog, or neither. Then,

    • if a dog is detected in the image, return the predicted breed.
    • if a human is detected in the image, return the resembling dog breed.
    • if neither is detected in the image, provide output that indicates an error.

    You are welcome to write your own functions for detecting humans and dogs in images, but feel free to use the face_detector and dog_detector functions developed above. You are required to use your CNN from Step 5 to predict dog breed.

    Some sample output for our algorithm is provided below, but feel free to design your own user experience!

    Sample Human Output

    (IMPLEMENTATION) Write your Algorithm

    In [60]:
    ### TODO: Write your algorithm.
    ### Feel free to use as many code cells as needed.
    def dog_human_dog_NA_detector(img_path):
        img = cv2.imread(img_path)
        cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
        return cv_rgb
        
    fig = plt.figure(figsize=(20, 8))
    for i in range(0,8):
        img_path_n=train_files[i]
        dog_breed = predict_breed_for_dog(img_path_n) 
        ax = fig.add_subplot(2, 4, i + 1, xticks=[], yticks=[])
        ax.imshow(dog_human_dog_NA_detector(img_path_n))
        ax.set_title("{} {} ({})".format(i,("Dog Breed :" if dog_detector(img_path_n) else "Human Looks Like :" if  face_detector(img_path_n) 
                                                                           else "Error "),dog_breed),color=("green" if dog_detector(img_path_n)
                                                                           else "red" if  face_detector(img_path_n)
                                                                           else "black"))
    

    Compare Predicted to True

    In [46]:
    def dog_breed_image_read(img_path):
        breed = predict_breed_for_dog(img_path) 
        
        # Display the image
        img = cv2.imread(img_path)
        cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    #    plt.imshow(cv_rgb)
    #    plt.show()
        return cv_rgb
    #print(dog_breed_image_read(test_files))
    #print("\n",dog_names)
    #predict_breed_for_dog(test_files)
    fig = plt.figure(figsize=(20, 8))
    
    for i in range(0,8):
        #print(dog_names [np.argmax(test_targets[i])]) #true_idx
        #print((test_files[i]))
        #print(predict_breed_for_dog(test_files[i])) #pred_idx
        ax = fig.add_subplot(2, 4, i + 1, xticks=[], yticks=[])
        ax.imshow(dog_breed_image_read(test_files[i]))
        pred_idx = predict_breed_for_dog(test_files[i])
        true_idx = dog_names [np.argmax(test_targets[i])]
        #print(pred_idx)
        #print(true_idx)
        ax.set_title("{} {} ({})".format(i,pred_idx, true_idx),color=("green" if pred_idx == true_idx else "red"))
        
    

    Step 7: Test Your Algorithm

    In this section, you will take your new algorithm for a spin! What kind of dog does the algorithm think that you look like? If you have a dog, does it predict your dog's breed accurately? If you have a cat, does it mistakenly think that your cat is a dog?

    (IMPLEMENTATION) Test Your Algorithm on Sample Images!

    Test your algorithm at least six images on your computer. Feel free to use any images you like. Use at least two human and two dog images.

    Question 6: Is the output better than you expected :) ? Or worse :( ? Provide at least three possible points of improvement for your algorithm.

    Answer: The below results on my 7 sample images is as per my expectations and have correctly identified however if you see the above training results there is one error in the test data sets where it has identified wrongly Belgian Sheep dogs as Ginat Schnazuer. These are very similar breeds and my program could not clearly identify the same.


    Items I would consider to improve algorithim will be (mainly to handle dual images next to each other, angular shift -- dog lying down horizontally and only portion of face visible, image pixcel distortation etc)

  • Increase the depth of layer

  • shear_range: Float. Shear Intensity (Shear angle in counter-clockwise direction as radians)

  • zoom_range: Float or [lower, upper]. Range for random zoom. If a float, [lower, upper] = [1-zoom_range, 1+zoom_range].

  • channel_shift_range: Float. Range for random channel shifts.

  • horizontal_flip: Boolean. Randomly flip inputs horizontally

  • In [63]:
    ## TODO: Execute your algorithm from Step 6 on
    ## at least 6 images on your computer.
    ## Feel free to use as many code cells as needed.
    image_sample_files = np.array(glob("samples/*"))
    print(image_sample_files)
    
    ['samples\\AllenGreenspan.jpg' 'samples\\Brittany_02625.jpg'
     'samples\\Curly-coated_retriever_03896.jpg' 'samples\\Diana.jpg'
     'samples\\Hullo0.JPG' 'samples\\Labrador_retriever_06449.jpg'
     'samples\\lion.jpeg']
    
    In [64]:
    fig = plt.figure(figsize=(20, 8))
    for i,image_sample_file  in enumerate(image_sample_files):
        img_path_n=image_sample_file
        dog_breed = predict_breed_for_dog(img_path_n) 
        ax = fig.add_subplot(2, 4, i + 1, xticks=[], yticks=[])
        ax.imshow(dog_human_dog_NA_detector(img_path_n))
        ax.set_title("{} {} ({})".format(i,("Dog Breed :" if dog_detector(img_path_n) else "Human Looks Like :" if  face_detector(img_path_n) 
                                                                           else "Neither but close match to dog breed:"),dog_breed),color=("green" if dog_detector(img_path_n)
                                                                           else "red" if  face_detector(img_path_n)
                                                                           else "black"))
    
    In [ ]: